# rule-engine

A rule engine is a computing system that applies rules to data to produce outcomes, commonly using inference engines for condition evaluation and action execution.

This is one page of public article previews, not the complete archive. Follow Next page to continue. Summaries are not the original full articles.

## From Code Review Comments to Reuse in Production: An Industrial Study of Extracted Rules

DevFeed: [From Code Review Comments to Reuse in Production: An Industrial Study of Extracted Rules](<https://devfeed.tech/articles/from-code-review-comments-to-reuse-in-production-an-industrial-study-of-extracted-rules-145783.md>)

Original publisher: [Read original article](<https://arxiv.org/abs/2611.25918>)

Author: Mutsuha Asada, Yuhei Nakasaka, Keito Tabuchi, Shinpei Hayashi

Published: 2026-11-20T05:00:00Z

Content type: article

Language: en

Sources: [arXiv: Software Engineering](<https://devfeed.tech/sources/arxiv-software-engineering.md>)

Topics: [Code review](<https://devfeed.tech/topics/code-review.md>), [ai code review](<https://devfeed.tech/topics/ai-code-review.md>), [rule-engine](<https://devfeed.tech/topics/rule-engine.md>)

Tags: [code-review](<https://devfeed.tech/tags/code-review.md>), [cs-se](<https://devfeed.tech/tags/cs-se.md>), [industrial](<https://devfeed.tech/tags/industrial.md>), [production](<https://devfeed.tech/tags/production.md>), [reuse](<https://devfeed.tech/tags/reuse.md>), [rules](<https://devfeed.tech/tags/rules.md>), [study](<https://devfeed.tech/tags/study.md>)

### AI overview

In a 65-day production study of Kleene, counting how often later review comments cited an extracted rule did not reveal what the rule contributed. Among 56 clauses from 12 rules cited as reasons, 2 were explicit in their source review threads, 38 added conditions, and 16 had no located source passage; many rule-attributed comments requested no change. On 20 constructed changes, a model using extracted rules met detection criteria on all 10 cases requiring a report, versus 8 of 10 using the unedited source threads, though the confidence interval included no difference. Of 464 deployed rules, 55.8% were named, 42.0% produced a non-withheld comment, and 20.3% led to a recorded code change; the largest drop came before a rule was named. Ranking rules in advance also showed no clear advantage over random selection. The authors recommend tracking each stage separately and evaluating rules against their source threads.

### Source excerpt

arXiv:2611.25918v1 Announce Type: new Abstract: Much organization-specific code review knowledge remains embedded in comments on individual changes, and recent systems reuse it by turning those comments into rules a language model applies to later changes. Teams must then decide which rules to keep, and the evidence at hand is how often the system's later comments name each rule. Tracing the rules of one such system, Kleene, through 65 days of production use, we found that such a count does not show what a rule contributed. A rule attribution indicates neither whether the source content was preserved nor whether the comment requested a change. Of 56 clauses in 12 rules that later comments named as their reason, 2 were explicit in a source review thread, 38 added conditions to a corresponding passage, and 16 had no passage we could locate; 45 of 79 further comments naming those rules requested no change, 27 of them while stating that the rule applied. On 20 constructed code changes, a reviewer model given only the unedited source thread met the detection criteria on 8 of the 10 that require a report, against 10 of 10 with the extracted rule, a difference whose confidence interval includes zero. Stage counts locate where rules drop out but not why: of 464 rules deployed while the rule-based reviewer subagent ran in their repository, 259 (55.8%) were named, 195 (42.0%) produced a comment that was not withheld, and 94 (20.3%) one with a recorded code change; the largest reduction occurred before any rule was named. Nor did the deployment-time properties we tested settle the choice in advance: retaining the top-ranked half preserved 53.8% of the rules that later produced a non-withheld comment, against 49.2% at random, a difference whose confidence interval includes zero. Judging extracted rules therefore calls for recording each stage separately and comparing each rule against the threads it was extracted from.

## Tether: Verified Stateful Policy Engine for System-Level Monitoring of Agents

DevFeed: [Tether: Verified Stateful Policy Engine for System-Level Monitoring of Agents](<https://devfeed.tech/articles/tether-verified-stateful-policy-engine-for-system-level-monitoring-of-agents-145548.md>)

Original publisher: [Read original article](<https://arxiv.org/abs/2611.25599>)

Author: Meghana Aparna Sistla, Alaia Solko-Breslin, Somesh Jha, Aditya Kanade, Krishnamurthy Dj Dvijotham

Published: 2026-11-20T05:00:00Z

Content type: article

Language: en

Sources: [arXiv: Artificial Intelligence](<https://devfeed.tech/sources/arxiv-artificial-intelligence.md>), [arXiv: Cryptography and Security](<https://devfeed.tech/sources/arxiv-cryptography-and-security.md>)

Topics: [policy-as-code](<https://devfeed.tech/topics/policy-as-code.md>), [Access Control](<https://devfeed.tech/topics/access-control.md>), [Lean](<https://devfeed.tech/topics/lean.md>), [rule-engine](<https://devfeed.tech/topics/rule-engine.md>), [agentic-coding](<https://devfeed.tech/topics/agentic-coding.md>)

Tags: [access-control](<https://devfeed.tech/tags/access-control.md>), [adversarial](<https://devfeed.tech/tags/adversarial.md>), [agents](<https://devfeed.tech/tags/agents.md>), [autonomous-ai-agents](<https://devfeed.tech/tags/autonomous-ai-agents.md>), [benchmarks](<https://devfeed.tech/tags/benchmarks.md>), [blocks](<https://devfeed.tech/tags/blocks.md>), [coding-agent](<https://devfeed.tech/tags/coding-agent.md>), [coding-agent-benchmarks](<https://devfeed.tech/tags/coding-agent-benchmarks.md>), [control](<https://devfeed.tech/tags/control.md>), [cs-ai](<https://devfeed.tech/tags/cs-ai.md>), [cs-cr](<https://devfeed.tech/tags/cs-cr.md>), [defense](<https://devfeed.tech/tags/defense.md>), [dsl](<https://devfeed.tech/tags/dsl.md>), [effects](<https://devfeed.tech/tags/effects.md>), [embedded](<https://devfeed.tech/tags/embedded.md>), [ephemeral](<https://devfeed.tech/tags/ephemeral.md>), [execution](<https://devfeed.tech/tags/execution.md>), [monitoring](<https://devfeed.tech/tags/monitoring.md>), [policy](<https://devfeed.tech/tags/policy.md>), [policy-engine](<https://devfeed.tech/tags/policy-engine.md>), [stateful](<https://devfeed.tech/tags/stateful.md>), [system](<https://devfeed.tech/tags/system.md>)

### AI overview

Tether checks agents' proposed actions by running each in an ephemeral isolated sandbox, inferring semantics from observed system calls, and enforcing stateful policies before allowing the action to proceed. Its policies are written in an embedded Lean 4 DSL for information flow, access control, and network security; formal verification can check global invariants such as blocking network access after a sensitive file is read. The abstract contrasts this with LLM monitors, which lack provable safety guarantees, and stateless enforcement, which can miss multi-step flows and side effects such as network calls inside scripts. On coding-agent benchmarks, one policy shared across tasks reportedly blocked 100% of in-scope attacks with a benign false-positive rate below 5%.

### Source excerpt

arXiv:2611.25599v1 Announce Type: cross Abstract: Autonomous AI agents are increasingly deployed for long-running tasks that execute arbitrary shell commands or programs. However, these actions can result in unauthorized network requests or persistent, malicious changes to the user's file system, processes, or privileges. Existing defense mechanisms rely on either LLM-based monitors, which lack provable safety guarantees, or stateless policy enforcement layers, which miss multi-step data flows and execution side-effects unmodeled by the policy language (e.g., network calls initiated inside a Python script). We introduce Tether, a framework for runtime enforcement of stateful security policies over the observed side-effects of an agent's actions. Tether executes each proposed action in an ephemeral, isolated sandbox, inferring abstract action semantics from observed system calls. Tether then enforces stateful policies over these inferred semantics, permitting actions to proceed only if they comply with the policy. Policies and action semantics are defined in an embedded DSL in Lean 4 capable of expressing diverse constraints, including information flow control (IFC), access control, and network security policies. Because actions can propagate labels from sources to destinations, a single rule can enforce fine-grained IFC policies across multi-step execution flows of arbitrary length. Expressing policies in Lean enables formal verification that they satisfy global security invariants (e.g., preventing network access after reading a sensitive file). We evaluate Tether on realistic coding-agent benchmarks across both benign and adversarial trajectories. With a single policy shared across all tasks, Tether blocks 100% of in-scope attacks while maintaining a benign false-positive rate under 5%.

## Design and Evaluation of Epira: An Automated Interview System for Case Investigation and Contact Tracing

DevFeed: [Design and Evaluation of Epira: An Automated Interview System for Case Investigation and Contact Tracing](<https://devfeed.tech/articles/design-and-evaluation-of-epira-an-automated-interview-system-for-case-investigation-and-contact-tracing-145072.md>)

Original publisher: [Read original article](<https://arxiv.org/abs/2611.25577>)

Author: Ishita Asnani, Shaghayegh Fazliani, Zachary Ryan, Kriyoshig Prasai, Chirag Kumar, Indra Gonzalez Ojeda, Marc Lipsitch, Madeleine Udell

Published: 2026-11-20T05:00:00Z

Content type: news

Language: en

Sources: [arXiv: Machine Learning](<https://devfeed.tech/sources/arxiv-machine-learning.md>)

Topics: [Large Language Model](<https://devfeed.tech/topics/llm.md>), [AI Chat](<https://devfeed.tech/topics/ai-chat.md>), [rule-engine](<https://devfeed.tech/topics/rule-engine.md>), [human review](<https://devfeed.tech/topics/human-review.md>), [Trustworthy AI](<https://devfeed.tech/topics/trustworthy-ai.md>), [human computer interaction](<https://devfeed.tech/topics/human-computer-interaction.md>), [trust](<https://devfeed.tech/topics/trust.md>)

Tags: [2](<https://devfeed.tech/tags/2.md>), [abstract](<https://devfeed.tech/tags/abstract.md>), [arxiv](<https://devfeed.tech/tags/arxiv.md>), [automated](<https://devfeed.tech/tags/automated.md>), [backends](<https://devfeed.tech/tags/backends.md>), [based](<https://devfeed.tech/tags/based.md>), [case](<https://devfeed.tech/tags/case.md>), [cases](<https://devfeed.tech/tags/cases.md>), [chatbots](<https://devfeed.tech/tags/chatbots.md>), [choice](<https://devfeed.tech/tags/choice.md>), [cloud](<https://devfeed.tech/tags/cloud.md>), [completeness](<https://devfeed.tech/tags/completeness.md>), [completion](<https://devfeed.tech/tags/completion.md>), [conclusion](<https://devfeed.tech/tags/conclusion.md>), [cs-hc](<https://devfeed.tech/tags/cs-hc.md>), [cs-lg](<https://devfeed.tech/tags/cs-lg.md>), [design](<https://devfeed.tech/tags/design.md>), [evaluation](<https://devfeed.tech/tags/evaluation.md>), [interview](<https://devfeed.tech/tags/interview.md>), [investigation](<https://devfeed.tech/tags/investigation.md>), [system](<https://devfeed.tech/tags/system.md>), [tracing](<https://devfeed.tech/tags/tracing.md>)

### AI overview

Epira combines LLM dialogue with a rule-based controller that enforces required questions, safety handoffs, and structured case records; new diseases can be added by uploading a questionnaire. In 108 simulated measles and gonorrhea cases, the local Qwen2.5 version captured 88.9% of required fields and 60.1% of known contacts. Open-ended interviewers captured 31.1-41.9 percentage points fewer required fields than the same models under Epira, and protocol control mattered more than model choice. In a separate study of 72 volunteers, completion was 100%, with mean usability of 82.95/100 and trust of 3.85/5. Local Qwen2.5 and cloud GPT-5.4 produced similar records. The authors support local use with human review of routed interviews and missed contacts; the results do not establish that all contacts are captured.

### Source excerpt

arXiv:2611.25577v1 Announce Type: new Abstract: Objective: Case investigation and contact tracing are labor-intensive. Web forms struggle when respondents need clarification before disclosing contacts, and open-ended chatbots may deviate from protocols. We developed and evaluated Epira, a protocol-controlled large language model (LLM) interviewer for multi-disease case investigation and contact tracing. Materials and Methods: Epira pairs LLM dialogue with a rule-based controller enforcing required questions, safety handoffs, and a structured case record, and adapts to new diseases by questionnaire upload. Local (Qwen2.5) and cloud (GPT-5.4) backends were tested. Study 1 used 108 simulated cases with ground-truth fields and contacts for measles and gonorrhea under Epira and open-ended conditions (864 sessions). Study 2 enrolled 72 volunteers in a 2$\times$2 disease-by-model interview assessing completion, usability, acceptability, and trust. Results: Across Study 1, Epira with a local LLM captured 88.9% of required fields and 60.1% of ground-truth contacts. Open-ended interviewers recovered 31.1-41.9 percentage points fewer required fields than the same models in Epira. In Study 2, completion was 100%; mean adapted System Usability Scale (SUS) was 82.95 $\pm$ 16.17 on a 0-100 scale and mean trust was 3.85 $\pm$ 0.77 on a 1-5 scale. Discussion: Protocol control drove completeness more than model choice. Local and cloud deployments produced similar records, supporting privacy-preserving local use with human review of routed interviews and missed contacts. Conclusion: Epira recovers required fields and contacts under diverse disclosure conditions, works with local and cloud LLMs, and received favorable usability and trust ratings.

## Stop Going in Circles:Loop-Triggered Routing for System 1 Decision Agents

DevFeed: [Stop Going in Circles:Loop-Triggered Routing for System 1 Decision Agents](<https://devfeed.tech/articles/stop-going-in-circles-loop-triggered-routing-for-system-1-decision-agents-145503.md>)

Original publisher: [Read original article](<https://arxiv.org/abs/2611.25363>)

Author: Zeming Yu, Huamin Ma, Yingyu Chen, Yuan Sun, Peng Hu, Yang Qin

Published: 2026-11-20T05:00:00Z

Content type: article

Language: en

Sources: [arXiv: Artificial Intelligence](<https://devfeed.tech/sources/arxiv-artificial-intelligence.md>)

Topics: [Artificial Intelligence](<https://devfeed.tech/topics/ai.md>), [llm-reasoning](<https://devfeed.tech/topics/llm-reasoning.md>), [rule-engine](<https://devfeed.tech/topics/rule-engine.md>)

Tags: [abstract](<https://devfeed.tech/tags/abstract.md>), [actions](<https://devfeed.tech/tags/actions.md>), [agent](<https://devfeed.tech/tags/agent.md>), [agents](<https://devfeed.tech/tags/agents.md>), [also](<https://devfeed.tech/tags/also.md>), [api](<https://devfeed.tech/tags/api.md>), [arxiv](<https://devfeed.tech/tags/arxiv.md>), [based](<https://devfeed.tech/tags/based.md>), [benchmarks](<https://devfeed.tech/tags/benchmarks.md>), [but](<https://devfeed.tech/tags/but.md>), [call](<https://devfeed.tech/tags/call.md>), [can](<https://devfeed.tech/tags/can.md>), [capabilities](<https://devfeed.tech/tags/capabilities.md>), [completion](<https://devfeed.tech/tags/completion.md>), [confidence](<https://devfeed.tech/tags/confidence.md>), [control](<https://devfeed.tech/tags/control.md>), [cs-ai](<https://devfeed.tech/tags/cs-ai.md>), [decision](<https://devfeed.tech/tags/decision.md>), [routing](<https://devfeed.tech/tags/routing.md>), [stop](<https://devfeed.tech/tags/stop.md>), [system](<https://devfeed.tech/tags/system.md>)

### AI overview

LoopRoute uses recurring state-action pairs as a signal that a System 1 decision agent has stalled. The decision model selects actions by default; when a loop is detected, an LLM provides brief, feedback-driven help, then control returns to the decision model. Later loops can trigger more assistance. The rule needs no extra training or model-specific confidence calibration. Across three benchmarks and seven LLMs, it improved task success over standalone Jev by mean gains of 21.0% on ALFWorld, 60.1% on MapReader, and 30.1% on TextWorld cooking. Compared with full LLM control, it substantially reduced mean cumulative API time and cost per task. The abstract also reports higher success rates for six frozen decision models, but provides no further methodological or quantitative detail for that result.

### Source excerpt

arXiv:2611.25363v1 Announce Type: new Abstract: System 1 decision models (represented by Jev) provide fast, low-cost action selection, but existing agent integrations do not reliably translate these local advantages into task completion. At certain points during execution, an agent may repeatedly select the same actions without making progress toward its goal. LLM-based policies offer strong reasoning capabilities that support planning within an agent, but invoking them at every step incurs substantial overhead. The effective coordination should therefore preserve Jev's fast local action selection while introducing LLM reasoning when execution stalls. We observe that recurring state-action pairs can signal stalled execution. We therefore propose LoopRoute, a simple but effective action router that uses this signal to allocate control within an agent. The decision model acts by default. When a loop is detected, the LLM provides brief, feedback-driven assistance and then returns control to the decision model. Later loops can request further assistance. The rule requires neither additional training nor model-specific confidence calibration. Across three benchmarks and seven LLMs, LoopRoute improves task success over standalone Jev with only limited use of the LLM policy. The mean gains are 21.0%, 60.1%, and 30.1% on ALFWorld, MapReader, and TextWorld cooking, respectively. Compared with full LLM control, LoopRoute also substantially reduces the mean cumulative API time and API-call cost per task. The same rule also improves the success rates of six frozen decision models.

## WorldRuler: Interactive World Modeling under Configurable Rules

DevFeed: [WorldRuler: Interactive World Modeling under Configurable Rules](<https://devfeed.tech/articles/worldruler-interactive-world-modeling-under-configurable-rules-142841.md>)

Original publisher: [Read original article](<https://arxiv.org/abs/2611.24869>)

Author: Mingliang Zhai, Chuanhao Li, Jiaming Tan, Zhen Li, Yuwei Wu, Yunde Jia, Zian Meng, Yuanyang Yin, Kaipeng Zhang

Published: 2026-11-19T05:00:00Z

Content type: article

Language: en

Sources: [arXiv: Artificial Intelligence](<https://devfeed.tech/sources/arxiv-artificial-intelligence.md>)

Topics: [AI for interactive applications](<https://devfeed.tech/topics/ai-for-interactive-applications.md>), [World models](<https://devfeed.tech/topics/world-models.md>), [rule-engine](<https://devfeed.tech/topics/rule-engine.md>)

Tags: [3d](<https://devfeed.tech/tags/3d.md>), [actions](<https://devfeed.tech/tags/actions.md>), [annotation](<https://devfeed.tech/tags/annotation.md>), [annotations](<https://devfeed.tech/tags/annotations.md>), [appearance](<https://devfeed.tech/tags/appearance.md>), [automated](<https://devfeed.tech/tags/automated.md>), [behaviors](<https://devfeed.tech/tags/behaviors.md>), [challenges](<https://devfeed.tech/tags/challenges.md>), [change](<https://devfeed.tech/tags/change.md>), [changes](<https://devfeed.tech/tags/changes.md>), [collection](<https://devfeed.tech/tags/collection.md>), [consistency](<https://devfeed.tech/tags/consistency.md>), [content](<https://devfeed.tech/tags/content.md>), [control](<https://devfeed.tech/tags/control.md>), [cs-ai](<https://devfeed.tech/tags/cs-ai.md>), [customizable](<https://devfeed.tech/tags/customizable.md>), [data](<https://devfeed.tech/tags/data.md>), [interactive](<https://devfeed.tech/tags/interactive.md>), [modeling](<https://devfeed.tech/tags/modeling.md>), [rules](<https://devfeed.tech/tags/rules.md>), [world](<https://devfeed.tech/tags/world.md>)

### AI overview

WorldRuler aims to make generative interactive worlds follow configurable rules while preserving them across save and reload. It represents rules, dependencies, and object states in an executable 3D rule graph, then links rule execution, video generation, and perceptual feedback in a loop. Dynamic object binding grounds rule-defined entities in visual instances and uses stored states to constrain generation from their first appearance, including when an affected object was initially unseen. Object attribute state replacement and counterfactual reconstruction training target designated changes while limiting effects on unrelated content. The authors also introduce RuleVerse, built using GTA-V and Habitat, with detailed state and event annotations, including rules that take effect before affected objects become visible. Their abstract reports improved rule-state persistence and realization of specified outcomes.

### Source excerpt

arXiv:2611.24869v1 Announce Type: new Abstract: Rules govern how player actions change interactive worlds. Different rules create distinct gameplay experiences within similar scenes and action spaces. For customizable gameplay, generative world models must follow user-defined rules beyond real-world behaviors. However, realizing user-defined rules presents two challenges: Rule-State Persistence and Rule-to-Visual Realization. The former requires preserving rule definitions, dependencies, and world states across saving and loading, so that restored worlds follow the established rules. The latter requires accurately realizing rule-induced changes on designated objects and attributes, particularly when affected objects are unseen. In this paper, we propose WorldRuler, which organizes rule definitions, dependencies, and object states in an executable 3D rule graph and connects rule execution, video generation, and perceptual feedback in a closed loop. The graph supports joint saving and restoration of rules and world states, while dynamic object binding grounds rule-specified entities in visual instances and uses stored states to constrain generation from their first appearance. We further introduce object attribute state replacement and counterfactual reconstruction training, enabling accurate attribute control and compatible appearance and motion changes while limiting interference with unrelated content. We develop an automated data collection and annotation pipeline using GTA-V and Habitat to construct RuleVerse, a dataset of rule-driven interactions with detailed state and event annotations, including cases where rules take effect before affected objects become visible. Extensive experiments demonstrate the effectiveness of WorldRuler in preserving rule-state consistency across saving and loading and accurately realizing rule outcomes on designated objects and attributes.

## Grading Executable Legal Rules Against Independent Records: Transitional Provisions and Korean SME Status

DevFeed: [Grading Executable Legal Rules Against Independent Records: Transitional Provisions and Korean SME Status](<https://devfeed.tech/articles/grading-executable-legal-rules-against-independent-records-transitional-provisions-and-korean-sme-status-142748.md>)

Original publisher: [Read original article](<https://arxiv.org/abs/2611.24055>)

Author: Jinhyung Bae

Published: 2026-11-19T05:00:00Z

Content type: article

Language: en

Sources: [arXiv: Artificial Intelligence](<https://devfeed.tech/sources/arxiv-artificial-intelligence.md>)

Topics: [rule-engine](<https://devfeed.tech/topics/rule-engine.md>), [llm-reasoning](<https://devfeed.tech/topics/llm-reasoning.md>), [audit](<https://devfeed.tech/topics/audit.md>)

Tags: [abstract](<https://devfeed.tech/tags/abstract.md>), [all](<https://devfeed.tech/tags/all.md>), [also](<https://devfeed.tech/tags/also.md>), [arxiv](<https://devfeed.tech/tags/arxiv.md>), [better](<https://devfeed.tech/tags/better.md>), [can](<https://devfeed.tech/tags/can.md>), [cases](<https://devfeed.tech/tags/cases.md>), [changes](<https://devfeed.tech/tags/changes.md>), [comparison](<https://devfeed.tech/tags/comparison.md>), [condition](<https://devfeed.tech/tags/condition.md>), [construction](<https://devfeed.tech/tags/construction.md>), [cs-ai](<https://devfeed.tech/tags/cs-ai.md>), [cs-cl](<https://devfeed.tech/tags/cs-cl.md>), [disclosures](<https://devfeed.tech/tags/disclosures.md>), [enterprise](<https://devfeed.tech/tags/enterprise.md>), [independent](<https://devfeed.tech/tags/independent.md>), [korean](<https://devfeed.tech/tags/korean.md>), [legal](<https://devfeed.tech/tags/legal.md>), [records](<https://devfeed.tech/tags/records.md>), [rules](<https://devfeed.tech/tags/rules.md>), [sme](<https://devfeed.tech/tags/sme.md>), [status](<https://devfeed.tech/tags/status.md>)

### AI overview

LexTemporis evaluates executable legal rules against 10,901 firm-years of SME status disclosed by Korean listed companies, rather than answers based on a grader's interpretation of the law. One transitional provision raises agreement with disclosures from 83-86% to 89-90%, showing that provisions outside the main statutory text can affect date-specific outcomes. The disclosures support the author's interpretation of a contested provision in 130 of 189 firm-years where readings diverge. Four LLMs generated rules from different selections of the law; a pre-registered comparison found no condition best across all four. In exploratory analyses, the apparent benefit of supplying relevant provisions was smaller against actual records than against the author's reading for three models, underscoring how grading against a shared interpretation can overstate performance.

### Source excerpt

arXiv:2611.24055v1 Announce Type: cross Abstract: Laws are increasingly turned into programs that decide cases, such as whether a firm was a small or medium enterprise (SME) on a given date, and large language models (LLMs) now write them. The programs are usually graded against answers written by someone who read the same law, so a program can score well by sharing the grader's reading rather than by getting the law right. LexTemporis grades them instead against records of what actually happened: 10,901 firm-years of SME status disclosed by Korean listed companies, with every version of the governing rules. An appendix adds comparison cases from US income tax and German hospital minimum volumes. The date changes outcomes mainly through provisions outside the main statutory text: one transitional provision raises agreement with the disclosures from 83-86% to 89-90%. The disclosures also test the author's reading of a contested provision, favouring it in 130 of 189 firm-years where readings diverge. Four LLMs wrote the rules, given different selections of the law. A pre-registered comparison finds no condition better across all four models. In exploratory analyses, supplying the right provisions looks better graded against the author's reading than against the records: the advantage shrinks for three models (for one, to between a quarter and three-fifths of its size), and per firm-year it is several times larger where the author's reading is wrong, partly by construction, since there reproducing that reading more faithfully must lose. Grading legal programs against the grader's own reading can overstate how well they work.

## Action Recognition vs. Detection + Tracking: When to Use Which

DevFeed: [Action Recognition vs. Detection + Tracking: When to Use Which](<https://devfeed.tech/articles/action-recognition-vs-detection-tracking-when-to-use-which-143831.md>)

Original publisher: [Read original article](<https://roboflow.com/blog/when-to-use-action-recognition>)

Author: Erik Kokalj

Published: 2026-11-18T22:22:42Z

Content type: comparison

Language: en

Sources: [Roboflow](<https://devfeed.tech/sources/roboflow-blog.md>)

Topics: [Video Analytics](<https://devfeed.tech/topics/video-analytics.md>), [image recognition](<https://devfeed.tech/topics/image-recognition.md>), [Machine learning](<https://devfeed.tech/topics/machine-learning.md>), [rule-engine](<https://devfeed.tech/topics/rule-engine.md>)

Tags: [action-recognition](<https://devfeed.tech/tags/action-recognition.md>), [basketball](<https://devfeed.tech/tags/basketball.md>), [bytetrack](<https://devfeed.tech/tags/bytetrack.md>), [cosmos](<https://devfeed.tech/tags/cosmos.md>), [cosmos-3](<https://devfeed.tech/tags/cosmos-3.md>), [detection](<https://devfeed.tech/tags/detection.md>), [f1](<https://devfeed.tech/tags/f1.md>), [fine-tuning](<https://devfeed.tech/tags/fine-tuning.md>), [fps](<https://devfeed.tech/tags/fps.md>), [motion](<https://devfeed.tech/tags/motion.md>), [tracking](<https://devfeed.tech/tags/tracking.md>), [video](<https://devfeed.tech/tags/video.md>), [workout](<https://devfeed.tech/tags/workout.md>)

### AI overview

The best approach depends on what the video question asks. Action recognition is suited to brief, subtle actions that require several frames: in the F1 test, Cosmos 3 Edge initially missed a quarter-second jack lift at 4 fps, but at 8 fps it became the strongest pit-stop system. In basketball, it caught all five dribbles, while a detector-based dribble rule caught one; it learned made shots from 13 examples. Detection plus ByteTrack is useful for visible objects and locations: it tracked tire removal better than action recognition, but missed jack lifts because the car barely moved and the crew obscured the jack. For body-pose tasks, a pretrained RF-DETR keypoint model plus simple geometry counted workout reps without training; action recognition often identified the exercise but merged repeated movements into one segment. The excerpt recommends enough labeled clips for action recognition, higher sampling rates for actions lasting a second or less, and adding a counter for reps. Detection and tracking depend on reliable detections, and pose distinctions such as squats versus lunges need keypoints rather than plain boxes.

### Source excerpt

We built action recognition and detection or keypoints + tracking systems for six video tasks, from F1 pit stops to workout reps. Here is when each approach wins.

## Optimizing Kubernetes admission control: Where Kyverno's CEL admission cost actually goes

DevFeed: [Optimizing Kubernetes admission control: Where Kyverno's CEL admission cost actually goes](<https://devfeed.tech/articles/optimizing-kubernetes-admission-control-where-kyverno-s-cel-admission-cost-actually-goes-142441.md>)

Original publisher: [Read original article](<https://www.cncf.io/blog/2026/11/18/optimizing-kubernetes-admission-control-where-kyvernos-cel-admission-cost-actually-goes/>)

Author: Raunak Madan (Software Engineer Intern, Nirmata)

Published: 2026-11-18T12:30:00Z

Content type: article

Language: en

Sources: [Cloud Native Computing Foundation](<https://devfeed.tech/sources/cloud-native-computing-foundation.md>)

Topics: [Kyverno](<https://devfeed.tech/topics/kyverno.md>), [rule-engine](<https://devfeed.tech/topics/rule-engine.md>), [control-plane](<https://devfeed.tech/topics/control-plane.md>)

Tags: [admission](<https://devfeed.tech/tags/admission.md>), [blog](<https://devfeed.tech/tags/blog.md>), [cel](<https://devfeed.tech/tags/cel.md>), [cost](<https://devfeed.tech/tags/cost.md>), [kubernetes](<https://devfeed.tech/tags/kubernetes.md>), [kyverno](<https://devfeed.tech/tags/kyverno.md>)

### AI overview

Profiling found that ValidatingPolicy's CEL path spent 86.8% of its allocations converting the admitted object, because it repeated the conversion for every matched policy. Kyverno 1.19.2 moves that conversion to once per request while preserving each engine's existing request.object behavior; the MutatingPolicy engine's 64-policy benchmark fell from 258,943 to 52,689 allocations and 10.59 ms to 2.97 ms. Image verification received the same memory-saving change. In a 16-policy comparison measured on an earlier Kyverno version, Kyverno was slower than native ValidatingAdmissionPolicy with one policy but faster with 16; the comparison has not been rerun after the fix. Benchmarks isolate the engine on one machine and omit webhook network and etcd costs. Remaining work includes the per-request JSON round-trip, mutating-policy object threading, and unaddressed single-replica tail latency.

### Source excerpt

Kyverno's CEL (Common Expression Language)-based engines exist to evaluate policy expressions. So when we profiled the admission path, we expected to find the bulk of the cost in CEL evaluation. It was not there. About 86%...

## Retail Merchandising Principles for Web Designers and Developers: Display Design, Ranking Rules and Measurement

DevFeed: [Retail Merchandising Principles for Web Designers and Developers: Display Design, Ranking Rules and Measurement](<https://devfeed.tech/articles/retail-merchandising-principles-for-web-designers-and-developers-display-design-ranking-rules-and-measurement-141352.md>)

Original publisher: [Read original article](<https://www.sitepoint.com/retail-merchandising-principles-for-web-designers-and-developers-display-design-ranking-rules-and-measurement/>)

Author: Div John

Published: 2026-11-18T09:46:28Z

Content type: article

Language: en

Sources: [SitePoint](<https://devfeed.tech/sources/sitepoint.md>)

Topics: [web design](<https://devfeed.tech/topics/web-design.md>), [User interface design](<https://devfeed.tech/topics/ui-design.md>), [Software Engineering](<https://devfeed.tech/topics/software-engineering.md>), [rule-engine](<https://devfeed.tech/topics/rule-engine.md>), [human centric design](<https://devfeed.tech/topics/human-centric-design.md>), [information retrieval](<https://devfeed.tech/topics/information-retrieval.md>), [designer developer handoff](<https://devfeed.tech/topics/designer-developer-handoff.md>), [CI/CD](<https://devfeed.tech/topics/cicd.md>)

Tags: [accessibility](<https://devfeed.tech/tags/accessibility.md>), [accessible](<https://devfeed.tech/tags/accessible.md>), [badge](<https://devfeed.tech/tags/badge.md>), [basics](<https://devfeed.tech/tags/basics.md>), [behavior](<https://devfeed.tech/tags/behavior.md>), [brand](<https://devfeed.tech/tags/brand.md>), [campaign](<https://devfeed.tech/tags/campaign.md>), [card](<https://devfeed.tech/tags/card.md>), [change](<https://devfeed.tech/tags/change.md>), [design](<https://devfeed.tech/tags/design.md>), [designers](<https://devfeed.tech/tags/designers.md>), [developer-tools](<https://devfeed.tech/tags/developer-tools.md>), [developers](<https://devfeed.tech/tags/developers.md>), [display](<https://devfeed.tech/tags/display.md>), [measurement](<https://devfeed.tech/tags/measurement.md>), [principles](<https://devfeed.tech/tags/principles.md>), [ranking](<https://devfeed.tech/tags/ranking.md>), [retail](<https://devfeed.tech/tags/retail.md>), [rules](<https://devfeed.tech/tags/rules.md>), [web](<https://devfeed.tech/tags/web.md>), [web-designers](<https://devfeed.tech/tags/web-designers.md>)

### AI overview

E-commerce merchandising works best when product visibility, ranking, accessibility and operational upkeep are treated as explicit interface rules. The article recommends curating the first screen, limiting visual emphasis, and encoding pins, boosts, brand caps and stock demotion in testable ranking logic; its Node.js example is a teaching sketch that omits production concerns such as personalization and search relevance. Price and badge colors should meet WCAG contrast thresholds, and promotions need start and end dates plus checks to prevent stale content. For layout changes, it demonstrates a two-proportion conversion test: illustrative results rise from 3.00% to 3.90% (z=2.21), while cautioning that this is not a general guarantee and assumes a preset sample size. Test one change with randomized comparable users, and coordinate design, engineering and operations so the shipped display matches the plan.

### Source excerpt

null Continue reading Retail Merchandising Principles for Web Designers and Developers: Display Design, Ranking Rules and Measurement on SitePoint.

## Human-like Semantic Navigation for Autonomous Driving using Knowledge Representation and Large Language Models

DevFeed: [Human-like Semantic Navigation for Autonomous Driving using Knowledge Representation and Large Language Models](<https://devfeed.tech/articles/human-like-semantic-navigation-for-autonomous-driving-using-knowledge-representation-and-large-language-models-139741.md>)

Original publisher: [Read original article](<https://arxiv.org/abs/2505.16498>)

Author: Augusto Luis Ballardini, Miguel \\'Angel Sotelo

Published: 2026-11-18T05:00:00Z

Content type: article

Language: en

Sources: [arXiv: Artificial Intelligence](<https://devfeed.tech/sources/arxiv-artificial-intelligence.md>)

Topics: [autonomous vehicles](<https://devfeed.tech/topics/autonomous-vehicles.md>), [llm-reasoning](<https://devfeed.tech/topics/llm-reasoning.md>), [Artificial Intelligence](<https://devfeed.tech/topics/ai.md>), [Large Language Model](<https://devfeed.tech/topics/llm.md>), [база знаний](<https://devfeed.tech/topics/tag-c51cc4bf76e5.md>), [rule-engine](<https://devfeed.tech/topics/rule-engine.md>)

Tags: [abstract](<https://devfeed.tech/tags/abstract.md>), [arxiv](<https://devfeed.tech/tags/arxiv.md>), [automation](<https://devfeed.tech/tags/automation.md>), [autonomous](<https://devfeed.tech/tags/autonomous.md>), [autonomous-driving](<https://devfeed.tech/tags/autonomous-driving.md>), [autonomous-navigation](<https://devfeed.tech/tags/autonomous-navigation.md>), [autonomous-vehicles](<https://devfeed.tech/tags/autonomous-vehicles.md>), [based](<https://devfeed.tech/tags/based.md>), [cs-ai](<https://devfeed.tech/tags/cs-ai.md>), [cs-ro](<https://devfeed.tech/tags/cs-ro.md>), [driving](<https://devfeed.tech/tags/driving.md>), [knowledge](<https://devfeed.tech/tags/knowledge.md>), [language](<https://devfeed.tech/tags/language.md>), [language-models](<https://devfeed.tech/tags/language-models.md>), [large](<https://devfeed.tech/tags/large.md>), [large-language-models](<https://devfeed.tech/tags/large-language-models.md>), [like](<https://devfeed.tech/tags/like.md>), [models](<https://devfeed.tech/tags/models.md>), [navigation](<https://devfeed.tech/tags/navigation.md>), [semantic](<https://devfeed.tech/tags/semantic.md>), [semantic-navigation](<https://devfeed.tech/tags/semantic-navigation.md>), [using](<https://devfeed.tech/tags/using.md>)

### AI overview

The method translates informal navigation instructions into Answer Set Programming (ASP) constraints using large language models, turning human-expressed intent into logic-based rules for autonomous driving. ASP's non-monotonic reasoning is intended to help vehicles adapt to changing road layouts, detours, or missing map data without relying on predefined maps. The abstract reports experimental evaluation and claims improved adaptability and explainability, but provides no quantitative results or implementation details.

### Source excerpt

arXiv:2505.16498v2 Announce Type: replace-cross Abstract: Achieving full automation in self-driving vehicles remains a challenge, especially in dynamic urban environments where navigation requires real-time adaptability. Existing systems struggle to handle navigation plans when faced with unpredictable changes in road layouts, spontaneous detours, or missing map data, due to their heavy reliance on predefined cartographic information. In this work, we explore the use of Large Language Models to generate Answer Set Programming rules by translating informal navigation instructions into structured, logic-based reasoning. ASP provides non-monotonic reasoning, allowing autonomous vehicles to adapt to evolving scenarios without relying on predefined maps. We present an experimental evaluation in which LLMs generate ASP constraints that encode real-world urban driving logic into a formal knowledge representation. By automating the translation of informal navigation instructions into logical rules, our method improves adaptability and explainability in autonomous navigation. Results show that LLM-driven ASP rule generation supports semantic-based decision-making, offering an explainable framework for dynamic navigation planning that aligns closely with how humans communicate navigational intent.

## Uber's Legal AI Agent Cuts Contract Review Time by 20% with Rules and RAG

DevFeed: [Uber's Legal AI Agent Cuts Contract Review Time by 20% with Rules and RAG](<https://devfeed.tech/articles/uber-s-legal-ai-agent-cuts-contract-review-time-by-20-with-rules-and-rag-135413.md>)

Original publisher: [Read original article](<https://www.infoq.com/news/2026/11/uber-legal-redlining-agent/>)

Author: Leela Kumili

Published: 2026-11-16T14:37:00Z

Content type: news

Language: en

Sources: [InfoQ](<https://devfeed.tech/sources/infoq.md>)

Topics: [AI Agent](<https://devfeed.tech/topics/ai-agent.md>), [human review](<https://devfeed.tech/topics/human-review.md>), [rule-engine](<https://devfeed.tech/topics/rule-engine.md>), [Agent Skills](<https://devfeed.tech/topics/agent-skills.md>)

Tags: [agent](<https://devfeed.tech/tags/agent.md>), [agentic](<https://devfeed.tech/tags/agentic.md>), [agentic-ai-architecture](<https://devfeed.tech/tags/agentic-ai-architecture.md>), [ai](<https://devfeed.tech/tags/ai.md>), [ai-agent](<https://devfeed.tech/tags/ai-agent.md>), [ai-ml-data-engineering](<https://devfeed.tech/tags/ai-ml-data-engineering.md>), [architecture-design](<https://devfeed.tech/tags/architecture-design.md>), [artificial-intelligence](<https://devfeed.tech/tags/artificial-intelligence.md>), [changes](<https://devfeed.tech/tags/changes.md>), [claude](<https://devfeed.tech/tags/claude.md>), [contract-review](<https://devfeed.tech/tags/contract-review.md>), [development](<https://devfeed.tech/tags/development.md>), [feedback](<https://devfeed.tech/tags/feedback.md>), [graphs](<https://devfeed.tech/tags/graphs.md>), [knowledge](<https://devfeed.tech/tags/knowledge.md>), [knowledge-graphs](<https://devfeed.tech/tags/knowledge-graphs.md>), [langgraph](<https://devfeed.tech/tags/langgraph.md>), [legal](<https://devfeed.tech/tags/legal.md>), [legal-ai](<https://devfeed.tech/tags/legal-ai.md>), [legal-matters](<https://devfeed.tech/tags/legal-matters.md>), [microsoft](<https://devfeed.tech/tags/microsoft.md>), [microsoft-office](<https://devfeed.tech/tags/microsoft-office.md>), [microsoft-word](<https://devfeed.tech/tags/microsoft-word.md>), [ml-data-engineering](<https://devfeed.tech/tags/ml-data-engineering.md>), [news](<https://devfeed.tech/tags/news.md>), [opensearch](<https://devfeed.tech/tags/opensearch.md>), [rag](<https://devfeed.tech/tags/rag.md>), [redis](<https://devfeed.tech/tags/redis.md>), [retrieval-augmented-generation](<https://devfeed.tech/tags/retrieval-augmented-generation.md>), [review](<https://devfeed.tech/tags/review.md>), [rules](<https://devfeed.tech/tags/rules.md>), [software-engineering](<https://devfeed.tech/tags/software-engineering.md>), [time](<https://devfeed.tech/tags/time.md>), [uber](<https://devfeed.tech/tags/uber.md>), [uber-legal-redlining-agent](<https://devfeed.tech/tags/uber-legal-redlining-agent.md>)

### AI overview

Uber's Legal Redlining Agent reports reducing average contract review time by more than 20%, with 91% accuracy in AI-generated decisions; lawyers still review and approve its recommendations. Integrated into Microsoft Word, it detects counterparties' edits, assesses risk and policy fit, and proposes clause changes with explanatory comments. Its RAG system combines legal playbooks with lawyer decisions, reasoning, and edited text. Similarity search and metadata filters narrow thousands of past interactions to about 20 candidates; a second model filters for relevance, while a configurable 365-day half-life gives more weight to recent examples. LangGraph coordinates parallel tasks, OpenSearch stores rules and feedback, and Redis holds session state. Deterministic rules override model suggestions for non-negotiable requirements, and a final check validates responses against retrieved rules. Early RAG struggled with retrieval accuracy, negotiation tone, and cases missing from examples, prompting the feedback loop. The excerpt also says Uber is exploring agentic harnesses, skills, legal ontologies, and knowledge graphs.

### Source excerpt

Uber's Legal Redlining Agent reviews contract changes in Microsoft Word using RAG, historical lawyer feedback, and deterministic policy rules. Built with LangGraph, OpenSearch, and Redis, the system reports more than 20% lower average contract review time and 91% accuracy in AI-generated decisions. Uber is exploring agentic harnesses, skills, legal ontologies, and knowledge graphs. By Leela Kumili

## Proof-Grounded Patient-Specific Clinical Explanations from Knowledge-Graph Reasoning

DevFeed: [Proof-Grounded Patient-Specific Clinical Explanations from Knowledge-Graph Reasoning](<https://devfeed.tech/articles/proof-grounded-patient-specific-clinical-explanations-from-knowledge-graph-reasoning-135013.md>)

Original publisher: [Read original article](<https://arxiv.org/abs/2610.06549>)

Author: Surajit Das

Published: 2026-11-16T05:00:00Z

Content type: article

Language: en

Sources: [arXiv: Artificial Intelligence](<https://devfeed.tech/sources/arxiv-artificial-intelligence.md>)

Topics: [Artificial Intelligence](<https://devfeed.tech/topics/ai.md>), [knowledge-engineering](<https://devfeed.tech/topics/knowledge-engineering.md>), [база знаний](<https://devfeed.tech/topics/tag-c51cc4bf76e5.md>), [rule-engine](<https://devfeed.tech/topics/rule-engine.md>), [audit](<https://devfeed.tech/topics/audit.md>), [Data Quality](<https://devfeed.tech/topics/data-quality.md>)

Tags: [audit](<https://devfeed.tech/tags/audit.md>), [auditability](<https://devfeed.tech/tags/auditability.md>), [availability](<https://devfeed.tech/tags/availability.md>), [clinical-decision-support](<https://devfeed.tech/tags/clinical-decision-support.md>), [consistency](<https://devfeed.tech/tags/consistency.md>), [cs-ai](<https://devfeed.tech/tags/cs-ai.md>), [implementation](<https://devfeed.tech/tags/implementation.md>), [knowledge-graph](<https://devfeed.tech/tags/knowledge-graph.md>), [provenance](<https://devfeed.tech/tags/provenance.md>), [reasoning](<https://devfeed.tech/tags/reasoning.md>), [trace](<https://devfeed.tech/tags/trace.md>)

### AI overview

The CKG Clinical Explanation Engine adds auditable, patient-specific traces downstream of a frozen, training-free clinical knowledge-graph reasoner. It converts inference states and disease knowledge into typed facts, explicit rule traces, provenance-linked conclusions, and policy-licensed recommendations. Separating evidence availability, representation completeness, and disease-specific activation preserves distinctions between observed zero-activation or partial evidence and unobserved evidence. Across six workbooks, several structural checks reached 100%; availability-representation consistency was 99.29%, with 12,117 disagreements. In paired comparisons over 90 traces, collapsing evidence states changed 20 action sets and removed 10 urgent-review occurrences; replay attributed all 40 action changes to reclassifying partial evidence as unobserved. These results support structural and implementation auditability, while clinical correctness and utility remain unestablished. Language-generation evaluation is future work.

### Source excerpt

arXiv:2610.06549v2 Announce Type: replace Abstract: Clinical decision support can lack auditable links from observations and encoded knowledge to conclusions and recommendations. Downstream of a frozen, training-free clinical knowledge-graph reasoner, the CKG Clinical Explanation Engine converts patient inference states and disease knowledge into typed facts, explicit rule traces, provenance-linked conclusions, and policy-licensed recommendations. Separate availability, representation completeness, and disease-specific activation distinguish observed zero-activation and partial evidence from unobserved evidence. Language-generation evaluation remains future work. Across six usable workbooks (11,289 patients, 33,867 patient-disease traces, 1,715,928 feature-evidence rows), IG-range validity, knowledge provenance, cross-sheet numerical fidelity (203,202/203,202), and exported fact/report field presence (33,867/33,867) reached 100.00%. Availability-representation consistency reached 99.29% (1,703,811/1,715,928), with 12,117 disagreements across five systematic feature-cohort patterns. Observed zero-activation and partial instances numbered 111,488 and 226,880, respectively. A separate seeded 30-patient end-to-end audit ran failure-free, passing all prespecified trace, licensing, provenance, and state-consistency checks. Paired comparisons (90 traces, fixed inherited diagnostic scores and policy catalogue) showed the availability-only baseline changed every report but no action sets. State collapse changed 20 action sets and removed 10 urgent-review occurrences. Rule replay reproduced all 40 action changes from reclassifying partial evidence as unobserved. The bounded validator detected all 450 single-fault mutants. Structural and implementation auditability were established. Clinical correctness and utility remain unestablished.

## Testing Conversation Preservation Across LLM APIs: A Rule-Based Oracle for Detecting Silent Information Loss

DevFeed: [Testing Conversation Preservation Across LLM APIs: A Rule-Based Oracle for Detecting Silent Information Loss](<https://devfeed.tech/articles/testing-conversation-preservation-across-llm-apis-a-rule-based-oracle-for-detecting-silent-information-loss-131717.md>)

Original publisher: [Read original article](<https://arxiv.org/abs/2611.15543>)

Author: Xin Liu, Xu Han, Yuxin Wang, Haoran Song, Mingjian Zhao, Lei Wang

Published: 2026-11-13T05:00:00Z

Content type: article

Language: en

Sources: [arXiv: Artificial Intelligence](<https://devfeed.tech/sources/arxiv-artificial-intelligence.md>), [arXiv: Software Engineering](<https://devfeed.tech/sources/arxiv-software-engineering.md>)

Topics: [open-responses](<https://devfeed.tech/topics/open-responses.md>), [rule-engine](<https://devfeed.tech/topics/rule-engine.md>), [Test-driven development](<https://devfeed.tech/topics/tdd.md>), [Large Language Model](<https://devfeed.tech/topics/llm.md>)

Tags: [agent](<https://devfeed.tech/tags/agent.md>), [apis](<https://devfeed.tech/tags/apis.md>), [compatibility](<https://devfeed.tech/tags/compatibility.md>), [controls](<https://devfeed.tech/tags/controls.md>), [conversations](<https://devfeed.tech/tags/conversations.md>), [cs-ai](<https://devfeed.tech/tags/cs-ai.md>), [cs-se](<https://devfeed.tech/tags/cs-se.md>), [error](<https://devfeed.tech/tags/error.md>), [evaluation](<https://devfeed.tech/tags/evaluation.md>), [exceptions](<https://devfeed.tech/tags/exceptions.md>), [experiments](<https://devfeed.tech/tags/experiments.md>), [fault-injection](<https://devfeed.tech/tags/fault-injection.md>), [gateways](<https://devfeed.tech/tags/gateways.md>), [generation](<https://devfeed.tech/tags/generation.md>), [human-review](<https://devfeed.tech/tags/human-review.md>), [information](<https://devfeed.tech/tags/information.md>), [interoperability](<https://devfeed.tech/tags/interoperability.md>), [layers](<https://devfeed.tech/tags/layers.md>), [llm](<https://devfeed.tech/tags/llm.md>), [llm-apis](<https://devfeed.tech/tags/llm-apis.md>), [oracle](<https://devfeed.tech/tags/oracle.md>), [testing](<https://devfeed.tech/tags/testing.md>)

### AI overview

The proposed isloss oracle tests whether LLM API translations preserve conversation information, catching losses that ordinary target schema checks can miss. It maps source and translated requests and replies to a shared representation, ConvIR, and uses explicit rules to separate legitimate format changes from avoidable loss and degradation due to target API limits. It needs neither a reference translation nor extra model calls. In evaluation, it detected 76% of planted faults with a 2% false alarm rate. Of 685 completed synthetic request translations, 18.5% passed static target checks while changing information the target could preserve; 36.4% of 558 translated reply streams caused exceptions or silent corruption in at least one tested official OpenAI SDK. Experiments across six models also found behavior changes when tasks depended on translated-away information.

### Source excerpt

arXiv:2611.15543v1 Announce Type: new Abstract: LLM applications increasingly rely on translators and compatibility layers to move conversations across APIs. Yet a translated request can be accepted by the target API while silently losing information about instruction roles, tool interactions, error states, or generation controls. We introduce isloss, a deterministic oracle for testing conversation-information preservation across LLM API translations. It maps source and translated requests and replies to a shared intermediate representation, ConvIR, and applies explicit rules to distinguish legitimate representation changes from information loss. Unlike schema validation or reference-based differential testing, isloss requires neither a reference translation nor additional model calls. It also distinguishes avoidable loss from degradation caused by target API limitations. We evaluate isloss using specification-based fault injection, human review, and experiments on translators, gateways, and client libraries. On held-out tests, it detects 95/125 planted faults (76.0%) with 2/100 false alarms (2.0%). Across a synthetic request suite spanning six configurations of five translator codebases, 127/685 completed translations (18.5%) pass static target checks while silently changing prompt, role, or tool information in ways the target API could have preserved. In translated replies, 203/558 streams (36.4%) cause exceptions or silent corruption in at least one tested official OpenAI SDK. Targeted experiments across six models show that translation-induced changes can alter model behaviour when tasks depend on the affected information. These results establish conversation preservation as a distinct correctness requirement for reliable LLM interoperability, agent execution, and model evaluation.

## Owned Edits to a Shared Decision Rule: Learning Private Exceptions as Typed, Gated, Withdrawable Weight Edits

DevFeed: [Owned Edits to a Shared Decision Rule: Learning Private Exceptions as Typed, Gated, Withdrawable Weight Edits](<https://devfeed.tech/articles/owned-edits-to-a-shared-decision-rule-learning-private-exceptions-as-typed-gated-withdrawable-weight-edits-133049.md>)

Original publisher: [Read original article](<https://arxiv.org/abs/2611.15917>)

Author: John J. O'Hare

Published: 2026-11-13T05:00:00Z

Content type: article

Language: en

Sources: [arXiv: Machine Learning](<https://devfeed.tech/sources/arxiv-machine-learning.md>)

Topics: [Machine learning](<https://devfeed.tech/topics/machine-learning.md>), [Federated Learning](<https://devfeed.tech/topics/federated-learning.md>), [Neural Network](<https://devfeed.tech/topics/neural-network.md>), [rule-engine](<https://devfeed.tech/topics/rule-engine.md>), [Large Language Model](<https://devfeed.tech/topics/llm.md>)

Tags: [2](<https://devfeed.tech/tags/2.md>), [about](<https://devfeed.tech/tags/about.md>), [abstract](<https://devfeed.tech/tags/abstract.md>), [acceptance](<https://devfeed.tech/tags/acceptance.md>), [all](<https://devfeed.tech/tags/all.md>), [apply](<https://devfeed.tech/tags/apply.md>), [arxiv](<https://devfeed.tech/tags/arxiv.md>), [base](<https://devfeed.tech/tags/base.md>), [bits](<https://devfeed.tech/tags/bits.md>), [body](<https://devfeed.tech/tags/body.md>), [can](<https://devfeed.tech/tags/can.md>), [cases](<https://devfeed.tech/tags/cases.md>), [change](<https://devfeed.tech/tags/change.md>), [changes](<https://devfeed.tech/tags/changes.md>), [cs-cy](<https://devfeed.tech/tags/cs-cy.md>), [cs-lg](<https://devfeed.tech/tags/cs-lg.md>), [decision](<https://devfeed.tech/tags/decision.md>), [edits](<https://devfeed.tech/tags/edits.md>), [exceptions](<https://devfeed.tech/tags/exceptions.md>), [learning](<https://devfeed.tech/tags/learning.md>), [private](<https://devfeed.tech/tags/private.md>), [weight](<https://devfeed.tech/tags/weight.md>)

### AI overview

The proposed mechanism lets each party add a private, withdrawable exception to a shared decision rule without changing decisions for other subjects. It compiles the published rule into a small ReLU network, then searches a small typed space of edits; an edit is proposed only when the party's own records pass a selection-aware evidence gate. In two preregistered synthetic analyses with 20 sampled hidden truths, at 30 records per party it scored 96.6%, versus 91.7% for the unchanged rule, and was within two points of seven tuned baselines. The same search without the structured prior or gate scored 84.0%. A third test found typed edits improved a shared-record-trained network by 5.4 points and edits over the published rule by 3.1. The authors say edits are recomputed from declarations, confined to units reserved for each party, exactly withdrawable, and disclose at most the information in the typed space. Reported weaknesses include shared truths, relaxations, ten-record settings, and a gate criterion that failed in both runs; a language model was about as accurate on one draw. These findings are reported for the described tests.

### Source excerpt

arXiv:2611.15917v1 Announce Type: new Abstract: Organisations that apply the same published decision rule often hold private evidence that the rule is wrong for their own cases, and will not pool it. We describe a learning mechanism for that setting whose unit of change is an owned edit. The published rule is compiled exactly into a small ReLU network, a structured prior. Each party's learner searches a small typed space of declarations, a few named values such as subject, scope, form and size, and proposes one only when its own records pass a selection-aware evidence gate. The learning is that gated search around the prior. In two analyses registered before they ran, on synthetic data with 20 sampled hidden truths, the learner at 30 records per party was non-inferior, within 2 points, to all seven tuned baselines (a confined gradient learner, Ditto, FedRep, federated averaging, two pooled networks and pooled gradient-boosted trees), scoring 96.6% against 91.7% for the unchanged rule; the same search without the prior or the gate scored 84.0%. In a third registered test, typed edits over a network trained on shared records beat that network alone by 5.4 points and the same edits over the published rule by 3.1. An acceptance body recomputes each edit from its declaration and writes it into units reserved for that party, so by construction an edit changes no decision about any other subject, withdraws exactly, discloses at most log2 of the typed space's size in bits, and can hide nothing beyond its declared values. We report where it loses: truths shared by all parties, relaxations, ten records, a pre-stated gate criterion that failed in both runs, and a language model given the records that decided about as accurately on one draw. On a rule-only base, a table of owned exceptions makes the same decisions.

## Mach2: Scientific Discovery Platform with Deterministic Execution Constraints

DevFeed: [Mach2: Scientific Discovery Platform with Deterministic Execution Constraints](<https://devfeed.tech/articles/mach2-scientific-discovery-platform-with-deterministic-execution-constraints-131695.md>)

Original publisher: [Read original article](<https://arxiv.org/abs/2611.14777>)

Author: Vansh Sharma, Venkat Raman

Published: 2026-11-13T05:00:00Z

Content type: article

Language: en

Sources: [arXiv: Artificial Intelligence](<https://devfeed.tech/sources/arxiv-artificial-intelligence.md>), [arXiv: Software Engineering](<https://devfeed.tech/sources/arxiv-software-engineering.md>)

Topics: [Agent Harness](<https://devfeed.tech/topics/agent-harness.md>), [Test-driven development](<https://devfeed.tech/topics/tdd.md>), [rule-engine](<https://devfeed.tech/topics/rule-engine.md>), [audit](<https://devfeed.tech/topics/audit.md>)

Tags: [abstract](<https://devfeed.tech/tags/abstract.md>), [acceptance](<https://devfeed.tech/tags/acceptance.md>), [acceptance-tests](<https://devfeed.tech/tags/acceptance-tests.md>), [agent](<https://devfeed.tech/tags/agent.md>), [agents](<https://devfeed.tech/tags/agents.md>), [also](<https://devfeed.tech/tags/also.md>), [analysis](<https://devfeed.tech/tags/analysis.md>), [artifacts](<https://devfeed.tech/tags/artifacts.md>), [arxiv](<https://devfeed.tech/tags/arxiv.md>), [can](<https://devfeed.tech/tags/can.md>), [changes](<https://devfeed.tech/tags/changes.md>), [compliance](<https://devfeed.tech/tags/compliance.md>), [continuous](<https://devfeed.tech/tags/continuous.md>), [contracts](<https://devfeed.tech/tags/contracts.md>), [coupling](<https://devfeed.tech/tags/coupling.md>), [cross](<https://devfeed.tech/tags/cross.md>), [cs-ai](<https://devfeed.tech/tags/cs-ai.md>), [cs-se](<https://devfeed.tech/tags/cs-se.md>), [discovery](<https://devfeed.tech/tags/discovery.md>), [execution](<https://devfeed.tech/tags/execution.md>), [physics-app-ph](<https://devfeed.tech/tags/physics-app-ph.md>), [platform](<https://devfeed.tech/tags/platform.md>), [with](<https://devfeed.tech/tags/with.md>)

### AI overview

Mach2 couples Eigen orchestration with deterministic checks of tool requests, source changes, acceptance tests, and artifact contracts. Requirements are executable predicates: mandatory failures block the checked operation, while diagnostic findings remain available for correction or review. User-specified TDD, CodeGrapher's abstract-syntax-tree analysis, and history-guided evolutionary search provide complementary verification and optimization. Task-specific agent packages can run through independently selected endpoints; the originating harness validates returned artifacts against declared interfaces and evidence requirements without requiring continuous supervision. Mach2 separately records model-internal-knowledge dependencies and proposes measures for rule compliance, engineering validity, and continuity across endpoints.

### Source excerpt

arXiv:2611.14777v1 Announce Type: new Abstract: Language-model agents can formulate workflows and invoke specialized tools, yet compliance with natural-language operating rules is not guaranteed by prompting alone. Mach2 addresses this limitation by coupling Eigen orchestration to deterministic, lint-like checks of tool requests, source changes, acceptance tests, and artifact contracts. Domain requirements are represented as executable predicates: mandatory failures prevent advancement of the checked operation, while diagnostic findings are retained for correction or review. User-specified test-driven development (TDD), CodeGrapher's abstract-syntax-tree analysis, and history-guided evolutionary search provide complementary verification and optimization mechanisms. Mach2 also generates task-specific agent packages for execution through independently selected endpoints. The originating harness validates returned artifacts against declared interfaces and evidence requirements without assuming continuous supervision of the remote run. Model-internal-knowledge (MIK) dependencies are recorded separately from source-grounded and test-supported findings. This paper formalizes the rule-gated execution model and the agent-package lifecycle, and proposes measurable criteria for evaluating rule compliance, engineering validity, and cross-endpoint task continuity.

## What's new in Kyverno 1.20

DevFeed: [What's new in Kyverno 1.20](<https://devfeed.tech/articles/what-s-new-in-kyverno-1-20-133514.md>)

Original publisher: [Read original article](<https://nirmata.com/2026/11/12/whats-new-in-kyverno-1-20/>)

Author: Jim Bugwadia

Published: 2026-11-12T15:00:00Z

Content type: release

Language: en

Sources: [Nirmata](<https://devfeed.tech/sources/nirmata.md>)

Topics: [Kyverno](<https://devfeed.tech/topics/kyverno.md>), [upgrade](<https://devfeed.tech/topics/upgrade.md>), [rule-engine](<https://devfeed.tech/topics/rule-engine.md>), [container image compliance](<https://devfeed.tech/topics/container-image-compliance.md>), [Helm charts](<https://devfeed.tech/topics/helm-charts.md>), [Test automation](<https://devfeed.tech/topics/test-automation.md>), [CI/CD](<https://devfeed.tech/topics/cicd.md>)

Tags: [2](<https://devfeed.tech/tags/2.md>), [3](<https://devfeed.tech/tags/3.md>), [about](<https://devfeed.tech/tags/about.md>), [absolute](<https://devfeed.tech/tags/absolute.md>), [admission](<https://devfeed.tech/tags/admission.md>), [admission-controller](<https://devfeed.tech/tags/admission-controller.md>), [after](<https://devfeed.tech/tags/after.md>), [ai](<https://devfeed.tech/tags/ai.md>), [cel](<https://devfeed.tech/tags/cel.md>), [changes](<https://devfeed.tech/tags/changes.md>), [cli](<https://devfeed.tech/tags/cli.md>), [engine](<https://devfeed.tech/tags/engine.md>), [exceptions](<https://devfeed.tech/tags/exceptions.md>), [image](<https://devfeed.tech/tags/image.md>), [image-verification](<https://devfeed.tech/tags/image-verification.md>), [kubernetes](<https://devfeed.tech/tags/kubernetes.md>), [kubernetes-security](<https://devfeed.tech/tags/kubernetes-security.md>), [kyverno](<https://devfeed.tech/tags/kyverno.md>), [news](<https://devfeed.tech/tags/news.md>), [policy](<https://devfeed.tech/tags/policy.md>), [policy-management](<https://devfeed.tech/tags/policy-management.md>), [upgrade](<https://devfeed.tech/tags/upgrade.md>), [verification](<https://devfeed.tech/tags/verification.md>), [what-s-new](<https://devfeed.tech/tags/what-s-new.md>)

### AI overview

Kyverno 1.20 removes execution of legacy policy types, including ClusterPolicy and CleanupPolicy, across controllers and the CLI. Existing resources remain but are marked not ready and unenforced; migrate them before upgrading. The Helm chart blocks upgrades when it finds legacy policies unless you acknowledge they will not be enforced. After upgrading, use `kyverno migrate` to update stored policy objects, and export legacy report history first because those results are pruned. The release also adds image verification for private Sigstore and GitHub attestations, CLI decision traces and OCI policy bundles, exception expiration and compensating controls for validating policies, and JSON mutation for object-root documents. Its CEL performance fix prepares each object once per request; reported engine benchmarks exclude webhook and API server latency. Kyverno 1.20 supports Kubernetes v1.34-v1.37; legacy CRDs are scheduled for removal in v1.21.

### Source excerpt

Kyverno 1.20 makes CEL the only policy engine and retires ClusterPolicy. See what changes for image verification, the CLI, exceptions and your upgrade.

## What is ABAC for AI agents?

DevFeed: [What is ABAC for AI agents?](<https://devfeed.tech/articles/what-is-abac-for-ai-agents-129765.md>)

Original publisher: [Read original article](<https://nirmata.com/2026/11/12/attribute-based-access-control-abac-for-ai-agents/>)

Author: Ritesh Patel

Published: 2026-11-12T04:23:18.277431Z

Content type: tutorial

Language: en

Sources: [Nirmata](<https://devfeed.tech/sources/nirmata.md>)

Topics: [Dynamic authorization rules](<https://devfeed.tech/topics/dynamic-authorization-rules.md>), [rule-engine](<https://devfeed.tech/topics/rule-engine.md>)

Tags: [2](<https://devfeed.tech/tags/2.md>), [2025](<https://devfeed.tech/tags/2025.md>), [2026](<https://devfeed.tech/tags/2026.md>), [3](<https://devfeed.tech/tags/3.md>), [403](<https://devfeed.tech/tags/403.md>), [about](<https://devfeed.tech/tags/about.md>), [access-control](<https://devfeed.tech/tags/access-control.md>), [access-controls](<https://devfeed.tech/tags/access-controls.md>), [access-tokens](<https://devfeed.tech/tags/access-tokens.md>), [after](<https://devfeed.tech/tags/after.md>), [agent](<https://devfeed.tech/tags/agent.md>), [agents](<https://devfeed.tech/tags/agents.md>), [ai](<https://devfeed.tech/tags/ai.md>), [ai-agents](<https://devfeed.tech/tags/ai-agents.md>), [ai-governance](<https://devfeed.tech/tags/ai-governance.md>), [environment](<https://devfeed.tech/tags/environment.md>), [here](<https://devfeed.tech/tags/here.md>), [how-to](<https://devfeed.tech/tags/how-to.md>), [map](<https://devfeed.tech/tags/map.md>), [rbac](<https://devfeed.tech/tags/rbac.md>), [security](<https://devfeed.tech/tags/security.md>), [tool](<https://devfeed.tech/tags/tool.md>), [what-is](<https://devfeed.tech/tags/what-is.md>)

### AI overview

ABAC evaluates an agent's request using attributes of the agent and person it represents, the tool and operation, and possibly the environment; unlike ACLs or RBAC, its rules can combine multiple attributes. For MCP tools, use claims from a token validated for the receiving server, define and assign authority for each attribute, and set freshness limits and failure behavior for environment lookups. Claude Code supports only narrow parameter rules: they match direct top-level input fields, with limits around omitted fields, multiple conditions, and primary content fields. Put broader checks in the server or gateway. The guide recommends testing rules, protecting decision logs, and weighing ABAC's maintenance cost against whether tools need complex attribute-based decisions.

### Source excerpt

ABAC grants or refuses an agent's action by checking attributes of the agent, the tool, the operation and the environment. Here is how to map it to tool calls.

## No finite rulebook can certify every safe program: an impossibility theorem for gatekeeping AI agents

DevFeed: [No finite rulebook can certify every safe program: an impossibility theorem for gatekeeping AI agents](<https://devfeed.tech/articles/no-finite-rulebook-can-certify-every-safe-program-an-impossibility-theorem-for-gatekeeping-ai-agents-128299.md>)

Original publisher: [Read original article](<https://arxiv.org/abs/2611.13367>)

Author: Tomas Perez-Acle

Published: 2026-11-11T05:00:00Z

Content type: article

Language: en

Sources: [arXiv: Artificial Intelligence](<https://devfeed.tech/sources/arxiv-artificial-intelligence.md>), [arXiv: Cryptography and Security](<https://devfeed.tech/sources/arxiv-cryptography-and-security.md>)

Topics: [Coq](<https://devfeed.tech/topics/coq.md>), [Local AI](<https://devfeed.tech/topics/local-ai.md>), [rule-engine](<https://devfeed.tech/topics/rule-engine.md>)

Tags: [adversarial](<https://devfeed.tech/tags/adversarial.md>), [ai](<https://devfeed.tech/tags/ai.md>), [computer-science](<https://devfeed.tech/tags/computer-science.md>), [cs-ai](<https://devfeed.tech/tags/cs-ai.md>), [cs-cr](<https://devfeed.tech/tags/cs-cr.md>), [data-privacy](<https://devfeed.tech/tags/data-privacy.md>), [gatekeeping](<https://devfeed.tech/tags/gatekeeping.md>), [machine](<https://devfeed.tech/tags/machine.md>), [simulations](<https://devfeed.tech/tags/simulations.md>)

### AI overview

The paper argues that no finite, machine-checkable pair of rulebooks can completely classify programs against rules such as never performing a forbidden action, keeping outputs independent of secrets, and eventually responding. A finite record can expose a violation of a prohibition, but cannot certify that a program will never violate it; for confidentiality and responsiveness, neither compliance nor violation can always be established this way. The authors characterize which rules about visible behavior can be confirmed and state that a sound rulebook of n bits omits a compliant program of about n bits. Time limits make checks finite but may require running nearly the full limit, which the abstract says is infeasible for a finite machine. Instead of deciding every case, a gatekeeper can provide a checkable list containing everything it must allow and nothing it must refuse. Reported simulations suggest the obstruction is adversarial rather than typical.

### Source excerpt

arXiv:2611.13367v1 Announce Type: new Abstract: An AI assistant writes small programs, and a gatekeeper must decide, from the code alone, whether each one may run. Suppose the rules are fixed: never perform a forbidden action, never let the output depend on a secret, always answer eventually. Can two rulebooks, one listing what is allowed and one listing what is not, both be complete, if every rule must be backed by evidence that a machine can check in finite time? We prove that they cannot. Checkable evidence can only establish what a finite record settles, and ``never'' and ``eventually'' are never settled by a finite record. For prohibitions the failure is one-sided: every violation can be caught by watching, but no safe program can be cleared for good. For confidentiality and responsiveness, neither side can be. We characterise exactly which rules about visible behaviour evidence can confirm, and we show that any sound rulebook of $n$ bits omits a compliant program of about $n$ bits. A time limit makes everything checkable, but only by running the program for essentially the whole limit, so a finite machine is not a feasible one. Finally, we show what a gatekeeper can still guarantee: not a verdict on every program, but a checkable list that contains everything it must allow and nothing it must refuse. Simulations on 2.2 million random programs and on self-referential adversaries show that the obstruction is adversarial rather than typical, and that the one error a longer test never removes is the liveness part of confidentiality.

## An Interpretable Fuzzy Tree for Turbofan Remaining Useful Life with an Auditable Per-Component Diagnostic Layer

DevFeed: [An Interpretable Fuzzy Tree for Turbofan Remaining Useful Life with an Auditable Per-Component Diagnostic Layer](<https://devfeed.tech/articles/an-interpretable-fuzzy-tree-for-turbofan-remaining-useful-life-with-an-auditable-per-component-diagnostic-layer-126150.md>)

Original publisher: [Read original article](<https://arxiv.org/abs/2611.11633>)

Author: Hugo Henry, Bharadwaj Dogga, Scott Philips, Kelly Cohen

Published: 2026-11-10T05:00:00Z

Content type: article

Language: en

Sources: [arXiv: Artificial Intelligence](<https://devfeed.tech/sources/arxiv-artificial-intelligence.md>)

Topics: [rule-engine](<https://devfeed.tech/topics/rule-engine.md>), [Neural Network](<https://devfeed.tech/topics/neural-network.md>)

Tags: [aggregation](<https://devfeed.tech/tags/aggregation.md>), [bounded](<https://devfeed.tech/tags/bounded.md>), [component](<https://devfeed.tech/tags/component.md>), [components](<https://devfeed.tech/tags/components.md>), [condition](<https://devfeed.tech/tags/condition.md>), [cs-ai](<https://devfeed.tech/tags/cs-ai.md>), [cycles](<https://devfeed.tech/tags/cycles.md>), [data](<https://devfeed.tech/tags/data.md>), [design](<https://devfeed.tech/tags/design.md>), [diagnostic](<https://devfeed.tech/tags/diagnostic.md>), [engine](<https://devfeed.tech/tags/engine.md>), [exhaustive](<https://devfeed.tech/tags/exhaustive.md>), [fewer](<https://devfeed.tech/tags/fewer.md>), [fleet](<https://devfeed.tech/tags/fleet.md>), [formal](<https://devfeed.tech/tags/formal.md>), [formal-verification](<https://devfeed.tech/tags/formal-verification.md>), [health](<https://devfeed.tech/tags/health.md>), [layer](<https://devfeed.tech/tags/layer.md>), [life](<https://devfeed.tech/tags/life.md>), [turbofan](<https://devfeed.tech/tags/turbofan.md>), [useful](<https://devfeed.tech/tags/useful.md>)

### AI overview

The proposed genetic fuzzy tree estimates turbofan remaining useful life while exposing rule-by-rule component diagnostics. It maps 13 cruise-condition sensor means to component health, bounded damage, spool damage, and RUL. Training fits and freezes health leaves to physical targets, applies a deterministic fuzzy damage map with fleet-referenced thresholds, and optimizes only the final aggregation for RUL. On the native N-CMAPSS split, it reports 8.44-cycle RMSE on the multi-fault subset and 11.51 across all modes, using 333 free parameters and 411 rules; compiled baselines have 32 to 12,000 times more rules. An exhaustive sensor search bounds diagnosability for this cruise-mean setup. Formal verification is not performed.

### Source excerpt

arXiv:2611.11633v1 Announce Type: new Abstract: Data-driven remaining useful life (RUL) estimators for turbofan engines are accurate but opaque, and their internal features typically encode remaining life rather than any specific component's state. This paper presents an interpretable genetic fuzzy tree that predicts RUL from cruise-condition sensor means on N-CMAPSS while exposing a per-component diagnostic layer inspectable rule by rule. The tree follows the engine hierarchy: sensors to component health, then to bounded component damage, spool damage, and RUL, so each level answers a different maintenance question. Training is staged so the prognostic objective cannot reshape the diagnostic layers: health leaves are fitted to physical targets and frozen, a deterministic fuzzy damage map with fleet-referenced thresholds is imposed on each component, and only the aggregation is optimized for RUL. It consumes 13 cruise-mean scalars per cycle rather than a trajectory, which removes any sequence model. On the native split it reaches 8.44 cycles RMSE on the multi-fault subset and 11.51 across all modes with 333 free parameters, 411 total rules, 32 to 12,000 times fewer than compiled baselines. An exhaustive sensor search bounds which components are diagnosable under the tested cruise-mean design, and the rule base is structured to admit formal verification, which this study does not discharge.

## TANSAR: Editable and Attributable Correlation Matrix Memories under Correlated Keys, with Rule-Constrained Multi-Hop Knowledge Editing

DevFeed: [TANSAR: Editable and Attributable Correlation Matrix Memories under Correlated Keys, with Rule-Constrained Multi-Hop Knowledge Editing](<https://devfeed.tech/articles/tansar-editable-and-attributable-correlation-matrix-memories-under-correlated-keys-with-rule-constrained-multi-hop-knowledge-editing-126270.md>)

Original publisher: [Read original article](<https://arxiv.org/abs/2611.12374>)

Author: Abdulhady Abas Abdulla, Jiyan Salim Mahmud, Fatemeh Daneshfar, Hussein Mohammed Ali, Mourad Oussalah, Muhammad Bilal

Published: 2026-11-10T05:00:00Z

Content type: article

Language: en

Sources: [arXiv: Artificial Intelligence](<https://devfeed.tech/sources/arxiv-artificial-intelligence.md>)

Topics: [AI Chat](<https://devfeed.tech/topics/ai-chat.md>), [knowledge-engineering](<https://devfeed.tech/topics/knowledge-engineering.md>), [rule-engine](<https://devfeed.tech/topics/rule-engine.md>), [llm-reasoning](<https://devfeed.tech/topics/llm-reasoning.md>), [infini-attention](<https://devfeed.tech/topics/infini-attention.md>)

Tags: [association](<https://devfeed.tech/tags/association.md>), [attribution](<https://devfeed.tech/tags/attribution.md>), [capacity](<https://devfeed.tech/tags/capacity.md>), [coherence](<https://devfeed.tech/tags/coherence.md>), [correlation](<https://devfeed.tech/tags/correlation.md>), [cs-ai](<https://devfeed.tech/tags/cs-ai.md>), [effect](<https://devfeed.tech/tags/effect.md>), [experiments](<https://devfeed.tech/tags/experiments.md>), [frame](<https://devfeed.tech/tags/frame.md>), [gate](<https://devfeed.tech/tags/gate.md>), [gpu](<https://devfeed.tech/tags/gpu.md>), [graph](<https://devfeed.tech/tags/graph.md>), [keys](<https://devfeed.tech/tags/keys.md>), [knowledge](<https://devfeed.tech/tags/knowledge.md>), [language](<https://devfeed.tech/tags/language.md>), [language-model](<https://devfeed.tech/tags/language-model.md>), [limit](<https://devfeed.tech/tags/limit.md>), [linear](<https://devfeed.tech/tags/linear.md>), [matrix](<https://devfeed.tech/tags/matrix.md>), [model](<https://devfeed.tech/tags/model.md>), [pruning](<https://devfeed.tech/tags/pruning.md>)

### AI overview

The paper analyzes Correlation Matrix Memories (CMMs) with correlated keys and introduces TANSAR for multi-hop knowledge editing. Its analysis gives an encoder-independent capacity bound, characterizes deletion's collateral effects while preserving certified associations, and finds that transparent attribution requires centred candidate keys. TANSAR traverses a directed acyclic association graph, uses sound ground Horn constraints to prune invalid paths, and sends a representation derived from the verified path into a frozen language model, with a Clopper-Pearson selective-risk gate. In reported experiments, Horn pruning lowered answer contradictions from 25.1% to 0.5%; Hopfield retrieval had 24% contradictions with stale associations retained and 0% when they were removed. Edit transactions took 13.6 ms on a standard CPU and 2.37 ms on GPU. The symbolic component matched gold labels on 600 ProofWriter theories and agreed with FOLIO at 94.0%.

### Source excerpt

arXiv:2611.12374v1 Announce Type: new Abstract: Correlation Matrix Memories (CMMs) store associations as sums of rank-one outer products, providing closed-form addressability, exact deletion, and linear attribution. While classical analyses assume random, orthogonal keys, semantic encoders produce correlated representations. We analyse CMMs under correlated keys using frame-theoretic coherence: (i) capacity bounds yielding an encoder-independent limit $n \le d(1+\varepsilon^2)$ for mean squared crosstalk $\varepsilon^2$; (ii) an exact characterisation of the collateral effect of deleting an association, guaranteeing that deletion never invalidates certified associations; and (iii) a geometric transparency bound demonstrating why attribution requires candidate sets of centred keys. On this basis, we introduce Topology-Aware Neuro-Symbolic Associative Retrieval (TANSAR). TANSAR traverses a directed acyclic association graph, prunes invalid paths via sound ground Horn constraints, and injects a representation derived strictly from the verified path into the residual stream of a frozen language model, guarded by a Clopper--Pearson selective-risk gate. Controlled experiments confirm all theoretical bounds. In multi-hop knowledge editing, Horn pruning reduces answer contradictions from 25.1% to 0.5%, whereas Hopfield retrieval incurs 24% contradictions when stale associations remain stored (and 0% if stale associations are removed). An end-to-end edit transaction requires 13.6 ms on a standard CPU and 2.37 ms on GPU. The symbolic component matches gold labels across all 600 ProofWriter theories and achieves 94.0% agreement on FOLIO, demonstrating sound and efficient multi-hop associative retrieval with exact unlearning.

## Auto-GlaKG: Automated Biomarker Reasoning for Explainable Glaucoma Detection

DevFeed: [Auto-GlaKG: Automated Biomarker Reasoning for Explainable Glaucoma Detection](<https://devfeed.tech/articles/auto-glakg-automated-biomarker-reasoning-for-explainable-glaucoma-detection-126092.md>)

Original publisher: [Read original article](<https://arxiv.org/abs/2611.11258>)

Author: Cheng Huang, Zeyu Han, Yian Hu, Yaopeng Ding, Jingxi Qiu, Yadi Liu

Published: 2026-11-10T05:00:00Z

Content type: article

Language: en

Sources: [arXiv: Artificial Intelligence](<https://devfeed.tech/sources/arxiv-artificial-intelligence.md>)

Topics: [Image](<https://devfeed.tech/topics/image.md>), [rule-engine](<https://devfeed.tech/topics/rule-engine.md>), [Algorithms](<https://devfeed.tech/topics/algorithms.md>)

Tags: [automated](<https://devfeed.tech/tags/automated.md>), [biomarkers](<https://devfeed.tech/tags/biomarkers.md>), [cdr](<https://devfeed.tech/tags/cdr.md>), [cross-validation](<https://devfeed.tech/tags/cross-validation.md>), [cs-ai](<https://devfeed.tech/tags/cs-ai.md>), [cs-cv](<https://devfeed.tech/tags/cs-cv.md>), [detection](<https://devfeed.tech/tags/detection.md>), [graph](<https://devfeed.tech/tags/graph.md>), [knowledge-graph](<https://devfeed.tech/tags/knowledge-graph.md>), [network](<https://devfeed.tech/tags/network.md>), [performance](<https://devfeed.tech/tags/performance.md>), [prediction](<https://devfeed.tech/tags/prediction.md>), [predictions](<https://devfeed.tech/tags/predictions.md>), [ratio](<https://devfeed.tech/tags/ratio.md>), [reasoning](<https://devfeed.tech/tags/reasoning.md>), [receiver](<https://devfeed.tech/tags/receiver.md>), [rules](<https://devfeed.tech/tags/rules.md>), [score](<https://devfeed.tech/tags/score.md>), [specificity](<https://devfeed.tech/tags/specificity.md>), [stage](<https://devfeed.tech/tags/stage.md>), [supervision](<https://devfeed.tech/tags/supervision.md>), [validation](<https://devfeed.tech/tags/validation.md>)

### AI overview

Auto-GlaKG links fundus-image predictions to glaucoma scores through explicit clinical-prior rules. A ResNet50 predicts cup-to-disc ratio (CDR) and six morphological signs; a knowledge-graph reasoner combines signed rule contributions. Across 689 images in stratified, group-aware five-fold cross-validation, it achieved pooled AUC 0.951 (95% CI 0.931-0.969), sensitivity 0.970, and specificity 0.779--matching the image classifier's AUC, with 0.2 ms added per image for reasoning. Substituting annotated CDR for predicted CDR raised fold-mean AUC from 0.953 to 0.995, highlighting CDR estimation as a priority for independent biomarker validation. These results describe the reported evaluation, not established performance in other settings.

### Source excerpt

arXiv:2611.11258v1 Announce Type: cross Abstract: Interpretable glaucoma detection requires a transparent connection between fundus images, clinically meaningful biomarkers, and diagnostic predictions. Knowledge graph (KG) reasoning provides this structure, but performance measured using annotated biomarkers does not establish performance when only images are available. We present Auto-GlaKG, an automated framework that combines image-based biomarker prediction with fixed clinical-prior rules. A ResNet50 network predicts the cup-to-disc ratio (CDR) and six morphological signs using biomarker supervision, and a KG reasoner aggregates their signed rule contributions into a glaucoma score. On 689 fundus images evaluated using stratified, group-aware five-fold cross-validation, Auto-GlaKG achieves a pooled area under the receiver operating characteristic curve (AUC) of 0.951 (95\% confidence interval: 0.931--0.969), with sensitivity of 0.970 and specificity of 0.779. The matched image classifier also achieves an AUC of 0.951, while the rule stage adds 0.2~ms per image. Replacing predicted CDR with its annotation increases fold-mean AUC from 0.953 to 0.995. The framework provides an explicit rule-level account of image-based predictions and identifies CDR estimation as a priority for validation using independently measured biomarkers.

## LoMASA: A Local Multi-Agent Sales Assistant with Confidence-Gated Delegation for B2B Quoting

DevFeed: [LoMASA: A Local Multi-Agent Sales Assistant with Confidence-Gated Delegation for B2B Quoting](<https://devfeed.tech/articles/lomasa-a-local-multi-agent-sales-assistant-with-confidence-gated-delegation-for-b2b-quoting-126089.md>)

Original publisher: [Read original article](<https://arxiv.org/abs/2611.11234>)

Author: Yu Liu, Xingbang Du

Published: 2026-11-10T05:00:00Z

Content type: article

Language: en

Sources: [arXiv: Artificial Intelligence](<https://devfeed.tech/sources/arxiv-artificial-intelligence.md>)

Topics: [Multi Agent Systems](<https://devfeed.tech/topics/multi-agent-systems.md>), [AI Chat](<https://devfeed.tech/topics/ai-chat.md>), [rule-engine](<https://devfeed.tech/topics/rule-engine.md>)

Tags: [agent](<https://devfeed.tech/tags/agent.md>), [b2b](<https://devfeed.tech/tags/b2b.md>), [confidence](<https://devfeed.tech/tags/confidence.md>), [cs-ai](<https://devfeed.tech/tags/cs-ai.md>), [cs-ma](<https://devfeed.tech/tags/cs-ma.md>), [delegation](<https://devfeed.tech/tags/delegation.md>), [local](<https://devfeed.tech/tags/local.md>), [multi-agent](<https://devfeed.tech/tags/multi-agent.md>), [sales](<https://devfeed.tech/tags/sales.md>)

### AI overview

LoMASA automates parts of B2B quote handling on a single desktop using two models: a 35B mixture-of-experts model selects decisions in one pass and uses answer-option probabilities for confidence, while a 27B dense model drafts customer replies and escalation briefs when confidence is low. Code calculates prices, margins, and options, and hard pricing rules cannot be overridden. In 500 simulated episodes (1,512 decisions), 38.0% of decisions were settled automatically, 42.5% were escalated by hard rules, and salespeople handled 19.5%; automatic decisions had a median latency of 2.67 seconds. On those decisions, accuracy matched a rule baseline (0.894), while below-floor margin commitments were less frequent (2.61% vs. 4.53%), at the cost of more manager escalations. Option probabilities were well calibrated without training, unlike the model's verbal confidence; the 27B model rarely changed decisions.

### Source excerpt

arXiv:2611.11234v1 Announce Type: new Abstract: Frontline sales staff spend much of their day re-coordinating quotes after customers change quantities, ask for discounts or new payment terms, or costs move. We present LoMASA (Local Multi-Agent Sales Assistant), a training-free multi-agent system that runs entirely on one desktop computer. A 35B mixture-of-experts model (System-1) makes each decision in a single forward pass and reads its confidence from the probabilities of the answer options. A 27B dense model (System-2) writes customer replies and escalation briefs and is consulted when System-1 is unsure. Code computes every price, margin and option, and hard pricing rules cannot be overridden. On a test split of 500 simulated episodes with 1512 decisions built from public supply-chain data, the salesperson has to decide only 19.5% of decisions: 38.0% are settled automatically, with a median latency of 2.67 seconds, and 42.5% are escalated to a manager by the hard rules. On the automatically settled decisions LoMASA is as accurate as a rule baseline (0.894 vs.\ 0.894) and makes fewer commitments whose margin later falls below the floor (2.61% vs 4.53%); the price is additional escalations to the manager. Option probabilities are well calibrated without any training, unlike confidence stated in words by the same model. We also report where the design does not help: the 27B model rarely changes a decision.

## Safeguarding Agentic AI in Radiology Workflows: A Multimodel Evaluation of Rule-based and Large Language Model-based Verification

DevFeed: [Safeguarding Agentic AI in Radiology Workflows: A Multimodel Evaluation of Rule-based and Large Language Model-based Verification](<https://devfeed.tech/articles/safeguarding-agentic-ai-in-radiology-workflows-a-multimodel-evaluation-of-rule-based-and-large-language-model-based-verification-126135.md>)

Original publisher: [Read original article](<https://arxiv.org/abs/2611.11541>)

Author: Semil Eminovic, Robin Schmidt, Bogdan Levita, Maximilian Lindholz, Tobias Penzkofer, Jawed Nawabi

Published: 2026-11-10T05:00:00Z

Content type: news

Language: en

Sources: [arXiv: Artificial Intelligence](<https://devfeed.tech/sources/arxiv-artificial-intelligence.md>)

Topics: [LLM evaluation / benchmarking](<https://devfeed.tech/topics/llm-evaluation-benchmarking.md>), [Large Language Model](<https://devfeed.tech/topics/llm.md>), [Hallucination detection](<https://devfeed.tech/topics/hallucination-detection.md>), [LLM security](<https://devfeed.tech/topics/llm-security.md>), [rule-engine](<https://devfeed.tech/topics/rule-engine.md>)

Tags: [2026](<https://devfeed.tech/tags/2026.md>), [agentic-ai](<https://devfeed.tech/tags/agentic-ai.md>), [claude-sonnet-5-5](<https://devfeed.tech/tags/claude-sonnet-5-5.md>), [cs-ai](<https://devfeed.tech/tags/cs-ai.md>), [cs-cl](<https://devfeed.tech/tags/cs-cl.md>), [cs-cy](<https://devfeed.tech/tags/cs-cy.md>), [documentation](<https://devfeed.tech/tags/documentation.md>), [error](<https://devfeed.tech/tags/error.md>), [evaluation](<https://devfeed.tech/tags/evaluation.md>), [gemma-4-31b](<https://devfeed.tech/tags/gemma-4-31b.md>), [glm-5-3](<https://devfeed.tech/tags/glm-5-3.md>), [gpt-6-1-sol](<https://devfeed.tech/tags/gpt-6-1-sol.md>), [large-language-model](<https://devfeed.tech/tags/large-language-model.md>), [llm](<https://devfeed.tech/tags/llm.md>), [minimax-m3](<https://devfeed.tech/tags/minimax-m3.md>), [permissions](<https://devfeed.tech/tags/permissions.md>), [qwen3-8-27b](<https://devfeed.tech/tags/qwen3-8-27b.md>), [risk](<https://devfeed.tech/tags/risk.md>), [rules](<https://devfeed.tech/tags/rules.md>), [simulation](<https://devfeed.tech/tags/simulation.md>), [verification](<https://devfeed.tech/tags/verification.md>), [workflows](<https://devfeed.tech/tags/workflows.md>)

### AI overview

In a simulation of 91 radiology documentation scenarios, safeguards reduced--but did not eliminate--high-severity errors by tool-using LLM agents. Across agents, violations occurred in 19.4% of episodes with permissions alone, 16.7% with deterministic rules, and 13.2% with rules plus a locally run LLM checker. The rules-only reduction was uncertain (95% CI included no change); adding the checker to rules reduced risk by 3.5 percentage points (95% CI: −6.3 to −0.7). Task success did not decrease detectably. In misleading-information variants, agents opened the altered document in 83% of episodes whether or not they erred, suggesting errors involved misjudgment rather than failure to access information. Most blocked actions would not have caused a violation and were sent back for revision. These findings come from synthetic, radiologist-reviewed scenarios and do not establish performance in clinical practice.

### Source excerpt

arXiv:2611.11541v1 Announce Type: new Abstract: Purpose: To estimate how often tool-using large language model (LLM) agents record clinically relevant false or unsupported information in simulated radiology documentation and how much safeguards applied before each action reduce these errors. Materials and Methods: In this simulation study (October 2026), 48 synthetic, radiologist-reviewed base scenarios covered finding communication, tumor board preparation and follow-up management. Forty-three also had a misleading-information variant (91 variants). Six LLM agents (Claude Sonnet 5.5, GPT-6.1-Sol, GLM-5.3, MiniMax M3, Gemma-4-31B, Qwen3.8-27B) completed each variant with permissions only, deterministic rules, and rules plus a locally run LLM checker (1638 episodes). The primary outcome was at least one executed high-severity violation per episode. Risk differences were estimated per agent at base-scenario level and, exploratorily, pooled across agents, with analyses of document access and blocks. Results: With permissions only, 7.3%-27.1% of episodes had a high-severity violation (pooled across agents, 19.4%). Pooled, it was 16.7% with rules (-2.8 percentage points; 95% CI: -6.1, 0.5) and 13.2% with rules plus the LLM checker (-6.2; 95% CI: -9.5, -3.0). Adding the LLM checker to rules changed it by -3.5 (95% CI: -6.3, -0.7). No difference between agents was detected (P = .43), and task success did not decrease detectably. In misleading variants, agents opened the altered document in 83% of episodes regardless of error, and 58% of erroneous statements cited it. Most blocked actions (64%-68%) would not have caused a violation and were returned for revision. Conclusion: Layered safeguards reduced but did not prevent high-severity errors. Errors after misleading information reflected misjudged rather than missed information.

## AutoPYara: Next-Gen YARA Rule Generator for Malware Family Clustering

DevFeed: [AutoPYara: Next-Gen YARA Rule Generator for Malware Family Clustering](<https://devfeed.tech/articles/autopyara-next-gen-yara-rule-generator-for-malware-family-clustering-122784.md>)

Original publisher: [Read original article](<https://arxiv.org/abs/2611.08608>)

Author: Mabon Ninan, Nhat Nguyen, Soumyajyoti Dutta, Sidharth Anil, Marcus Botacin

Published: 2026-11-09T05:00:00Z

Content type: article

Language: en

Sources: [arXiv: Cryptography and Security](<https://devfeed.tech/sources/arxiv-cryptography-and-security.md>), [arXiv: Machine Learning](<https://devfeed.tech/sources/arxiv-machine-learning.md>)

Topics: [clustering](<https://devfeed.tech/topics/clustering.md>), [Tool](<https://devfeed.tech/topics/tool.md>), [rule-engine](<https://devfeed.tech/topics/rule-engine.md>), [pip](<https://devfeed.tech/topics/pip.md>)

Tags: [abstract](<https://devfeed.tech/tags/abstract.md>), [automatic](<https://devfeed.tech/tags/automatic.md>), [cluster](<https://devfeed.tech/tags/cluster.md>), [clustering](<https://devfeed.tech/tags/clustering.md>), [coverage](<https://devfeed.tech/tags/coverage.md>), [cs-cr](<https://devfeed.tech/tags/cs-cr.md>), [cs-lg](<https://devfeed.tech/tags/cs-lg.md>), [data](<https://devfeed.tech/tags/data.md>), [datasets](<https://devfeed.tech/tags/datasets.md>), [diversity](<https://devfeed.tech/tags/diversity.md>), [effective](<https://devfeed.tech/tags/effective.md>), [generation](<https://devfeed.tech/tags/generation.md>), [generator](<https://devfeed.tech/tags/generator.md>), [hashing](<https://devfeed.tech/tags/hashing.md>), [implement](<https://devfeed.tech/tags/implement.md>), [incident-response](<https://devfeed.tech/tags/incident-response.md>), [malware](<https://devfeed.tech/tags/malware.md>), [next](<https://devfeed.tech/tags/next.md>), [open-source](<https://devfeed.tech/tags/open-source.md>), [package](<https://devfeed.tech/tags/package.md>), [pip-install](<https://devfeed.tech/tags/pip-install.md>), [yara](<https://devfeed.tech/tags/yara.md>)

### AI overview

AutoPYara targets degraded YARA rule quality caused by diverse malware within a family and noisy clustering. It selects cluster centroids using similarity hashing and estimates family subclusters with a heuristic based on historical VirusTotal data. The authors report 14 percentage points higher rule coverage, 7.1 percentage points lower average variance than AutoYara, and 10% better threat-hunting accuracy. They provide it as an open-source Python package installable with `pip install autopyara`. These are abstract-level reported results; the supplied excerpt gives no evaluation details.

### Source excerpt

arXiv:2611.08608v1 Announce Type: new Abstract: Millions of malware variants threaten users daily, such that clustering these samples is key for fast incident response. AutoYara is the current state-of-the-art for automatic YARA rule generation, but the task remains challenging on realistic datasets, where intra-family diversity and noisy clustering can degrade rule quality. We present AutoPYara, a solution armored with: (1) precise cluster centroid selection based on similarity hashing, and (2) an improved heuristic for estimating subclusters within a malware family using historical VirusTotal data. We implement AutoPYara as an open-source Python package (pip install autopyara) to support reproducibility and practical use. We show that AutoPYara, (1) increases rule coverage by 14 percentage points; (2) reduces variance by 7.1 percentage points, comparing the average variance AutoPYara exhibits to that of AutoYara; and (3) improves threat-hunting accuracy by 10%. AutoPYara produces robust and effective YARA rules than prior automatic approaches in practical settings.

[Next page](<https://devfeed.tech/topics/rule-engine.md?cursor=WyIyMDI2LTExLTA5VDA1OjAwOjAwKzAwOjAwIiwgIjhkZDQ5ZTg5LTFjYTYtNDUwMi04NTY5LTgyMmFmMzRhOWEwMCJd>)