# experience

Published articles for experience.

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

## Auto-Alpha: A Framework for Autonomous Factor Research

DevFeed: [Auto-Alpha: A Framework for Autonomous Factor Research](<https://devfeed.tech/articles/auto-alpha-a-framework-for-autonomous-factor-research-145578.md>)

Original publisher: [Read original article](<https://arxiv.org/abs/2611.25732>)

Author: Hui Cai, Kangcheng Deng, Tianyi Ma, Qixuan Yang, Jiacheng Lu, Rui Sun, Beidi Luan, Jing Li, Zuo Bai

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

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: [agentic workflows](<https://devfeed.tech/topics/agentic-workflows.md>)

Tags: [admission](<https://devfeed.tech/tags/admission.md>), [agent](<https://devfeed.tech/tags/agent.md>), [alpha](<https://devfeed.tech/tags/alpha.md>), [audit](<https://devfeed.tech/tags/audit.md>), [autonomous](<https://devfeed.tech/tags/autonomous.md>), [backtest](<https://devfeed.tech/tags/backtest.md>), [cs-ai](<https://devfeed.tech/tags/cs-ai.md>), [cs-se](<https://devfeed.tech/tags/cs-se.md>), [experience](<https://devfeed.tech/tags/experience.md>), [factor](<https://devfeed.tech/tags/factor.md>), [feedback](<https://devfeed.tech/tags/feedback.md>), [framework](<https://devfeed.tech/tags/framework.md>), [implementation](<https://devfeed.tech/tags/implementation.md>), [insights](<https://devfeed.tech/tags/insights.md>), [library](<https://devfeed.tech/tags/library.md>), [q-fin-cp](<https://devfeed.tech/tags/q-fin-cp.md>), [reports](<https://devfeed.tech/tags/reports.md>), [research](<https://devfeed.tech/tags/research.md>), [review](<https://devfeed.tech/tags/review.md>), [semantic](<https://devfeed.tech/tags/semantic.md>), [signal](<https://devfeed.tech/tags/signal.md>), [test](<https://devfeed.tech/tags/test.md>), [testing](<https://devfeed.tech/tags/testing.md>)

### AI overview

Auto-Alpha links report-guided factor discovery to iterative implementation and backtesting: an idea agent proposes economic mechanisms and signals, while a research agent refines candidates using backtest feedback. Semantic and quantitative reviews plus human admission govern entry into a persistent factor library; an Experience Library stores research insights. From 120 reports, the system built 32 initial factors, then produced 420 implementations and retained 62 more. In a fixed 105-factor snapshot, median 1-day Rank IC was 0.0301; a weekly Top-300 portfolio returned 16.81% annualized after stated costs over 2020-2026, compared with 1.71% for the CSI 1000 price index. These are retrospective results; controlled evaluation is needed to isolate component contributions.

### Source excerpt

arXiv:2611.25732v1 Announce Type: cross Abstract: Autonomous factor research requires generating economic hypotheses as well as implementing and testing them. We present Auto-Alpha, a framework that connects report-guided initialization to self-directed research through a test-driven implementation loop. An idea agent proposes economic mechanisms and signal constructions; a research agent develops them with backtest feedback. Semantic audit, quantitative review, and human admission select factors for a persistent library, while an Experience Library retains research insights. Starting from 120 research reports, Auto-Alpha builds an initial library of 32 factors. It then generates new ideas and refines existing candidates, producing 420 implementations and retaining 62 additional factors. We evaluate factor quality and portfolio performance using a fixed library snapshot of 105 factors. The median 1-day Rank IC is 0.0301, and a weekly Top-300 portfolio returns 16.81% annualized after stated costs over 2020-2026, versus 1.71% for the CSI 1000 price index. These retrospective results document an operating research process and its artifacts; controlled evaluation is needed to isolate the contribution of its components.

## FlashX: Scaling Robotic Experience Reuse for Embodied Agents with CBA-Based NAND Flash

DevFeed: [FlashX: Scaling Robotic Experience Reuse for Embodied Agents with CBA-Based NAND Flash](<https://devfeed.tech/articles/flashx-scaling-robotic-experience-reuse-for-embodied-agents-with-cba-based-nand-flash-145540.md>)

Original publisher: [Read original article](<https://arxiv.org/abs/2611.25580>)

Author: Yuseon Choi, Sanghyeon Park, Eunjong Lee, Hanseo Lee, Sangjin Kim

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

Content type: article

Language: en

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

Topics: [Humanoid Robots](<https://devfeed.tech/topics/humanoid-robots.md>), [Artificial Intelligence](<https://devfeed.tech/topics/ai.md>), [Machine learning](<https://devfeed.tech/topics/machine-learning.md>), [diffusion-transformers](<https://devfeed.tech/topics/diffusion-transformers.md>), [Cache](<https://devfeed.tech/topics/cache.md>), [High-Performance Computing](<https://devfeed.tech/topics/high-performance-computing.md>), [AI Infrastructure](<https://devfeed.tech/topics/ai-infrastructure.md>), [Distributed Systems & Parallel Computing](<https://devfeed.tech/topics/distributed-systems-parallel-computing.md>)

Tags: [abstract](<https://devfeed.tech/tags/abstract.md>), [acceleration](<https://devfeed.tech/tags/acceleration.md>), [acquisition](<https://devfeed.tech/tags/acquisition.md>), [actions](<https://devfeed.tech/tags/actions.md>), [agents](<https://devfeed.tech/tags/agents.md>), [architecture](<https://devfeed.tech/tags/architecture.md>), [array](<https://devfeed.tech/tags/array.md>), [arxiv](<https://devfeed.tech/tags/arxiv.md>), [bandwidth](<https://devfeed.tech/tags/bandwidth.md>), [banks](<https://devfeed.tech/tags/banks.md>), [based](<https://devfeed.tech/tags/based.md>), [better](<https://devfeed.tech/tags/better.md>), [cs-ai](<https://devfeed.tech/tags/cs-ai.md>), [cs-ar](<https://devfeed.tech/tags/cs-ar.md>), [cs-ro](<https://devfeed.tech/tags/cs-ro.md>), [experience](<https://devfeed.tech/tags/experience.md>), [flash](<https://devfeed.tech/tags/flash.md>), [nand](<https://devfeed.tech/tags/nand.md>), [nand-flash](<https://devfeed.tech/tags/nand-flash.md>), [reuse](<https://devfeed.tech/tags/reuse.md>), [robotic](<https://devfeed.tech/tags/robotic.md>), [scaling](<https://devfeed.tech/tags/scaling.md>), [with](<https://devfeed.tech/tags/with.md>)

### AI overview

FlashX uses CBA-based NAND flash to scale retrieval of successful robotic experiences that warm-start VLA and WAM action generation, aiming to improve adaptation to unfamiliar environments while reducing inference cost. It combines parallel pSLC key search using interleaved Compute-in-Flash engines with dense TLC latent storage; proprioception-aware clustering reduces search and supports incremental additions without global reclustering. Across the evaluated workloads, task success improved by up to 12.24 percentage points in the dense configuration and 7.43 points in the compute-efficient configuration. FlashX reports 2.57x the search throughput of the strongest prior NAND search architecture under pSLC sensing, with clustering reducing search a further 9.7x. Compared with SSD-based retrieval, it reports 1.29-3.95x end-to-end speedup and 1.12-2.34x energy efficiency improvement.

### Source excerpt

arXiv:2611.25580v1 Announce Type: cross Abstract: Embodied agents are emerging as a promising paradigm for complex robotic tasks, leveraging Vision-Language-Action (VLA) and World Action Models (WAMs) to generate actions through Diffusion Transformer (DiT)-based iterative denoising. However, existing models often struggle to generalize to out-of-distribution (OOD) environments. We observe that retrieving successful experiences and using them to warm-start denoising enables agents to better adapt to unfamiliar situations while reducing computational overhead. This creates a new opportunity to improve task accuracy and inference efficiency simultaneously, in contrast to conventional acceleration techniques that trade accuracy for computation. However, realizing this opportunity requires scaling experience banks, introducing substantial storage and search overhead that challenges the capacity, bandwidth, and power constraints of edge systems. We propose FlashX, a CMOS Directly Bonded to Array (CBA)-based NAND flash architecture for scalable experience reuse in embodied agents. FlashX exploits CBA's high internal bandwidth for parallel pSLC key search, employing interleaved Compute-in-Flash (CIF) engines to sustain sensing throughput while retaining dense TLC latent storage. Proprioception-aware clustering further reduces search overhead and enables incremental experience acquisition without global reclustering. Evaluations across VLA and WAM workloads demonstrate up to 12.24\%p and 7.43\%p improvements in task success under dense and compute-efficient configurations, respectively. FlashX achieves 2.57$\times$ higher search throughput than the strongest prior NAND search architecture under pSLC sensing, with a further 9.7$\times$ search reduction through clustering. This delivers 1.29--3.95$\times$ end-to-end speedup and 1.12--2.34$\times$ energy efficiency improvement over SSD-based retrieval.

## PlayCraft: Designing Engaging Games from Agent Play Experience

DevFeed: [PlayCraft: Designing Engaging Games from Agent Play Experience](<https://devfeed.tech/articles/playcraft-designing-engaging-games-from-agent-play-experience-145696.md>)

Original publisher: [Read original article](<https://arxiv.org/abs/2611.26624>)

Author: Weihao Cheng

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

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>), [Machine learning](<https://devfeed.tech/topics/machine-learning.md>), [Learning](<https://devfeed.tech/topics/learning.md>)

Tags: [agent](<https://devfeed.tech/tags/agent.md>), [ai](<https://devfeed.tech/tags/ai.md>), [automated](<https://devfeed.tech/tags/automated.md>), [cs-ai](<https://devfeed.tech/tags/cs-ai.md>), [cs-cv](<https://devfeed.tech/tags/cs-cv.md>), [designing](<https://devfeed.tech/tags/designing.md>), [engagement](<https://devfeed.tech/tags/engagement.md>), [experience](<https://devfeed.tech/tags/experience.md>), [framework](<https://devfeed.tech/tags/framework.md>), [game](<https://devfeed.tech/tags/game.md>), [games](<https://devfeed.tech/tags/games.md>), [generate](<https://devfeed.tech/tags/generate.md>), [intelligence](<https://devfeed.tech/tags/intelligence.md>), [learning](<https://devfeed.tech/tags/learning.md>), [measuring](<https://devfeed.tech/tags/measuring.md>), [memory](<https://devfeed.tech/tags/memory.md>), [performance](<https://devfeed.tech/tags/performance.md>), [play](<https://devfeed.tech/tags/play.md>), [predictions](<https://devfeed.tech/tags/predictions.md>), [progress](<https://devfeed.tech/tags/progress.md>), [reflections](<https://devfeed.tech/tags/reflections.md>), [reports](<https://devfeed.tech/tags/reports.md>)

### AI overview

PlayCraft revises games through repeated cycles of agent play and game crafting to support player mastery. A vision-based player agent completes repeated episodes with memory and reports its progress, outcomes, decisions, and reflections. A crafter uses those reports to estimate abilities, predict performance at different skill levels, compare progress with a target learning curve, and revise opportunities to improve decisions and execution. In fresh-run evaluations across three agent intelligence settings, PlayCraft produced larger learning gains and steeper learning slopes than the initial games in most cases, and beat the baselines on the macro-average of both metrics at every setting.

### Source excerpt

arXiv:2611.26624v1 Announce Type: new Abstract: Recent AI agents can generate playable games, but these games are not necessarily engaging. Mastery is a key source of engagement, yet designing for it requires extensive human play and tuning. We propose PlayCraft, an automated framework that uses a vision-based agent to play games and draws on its detailed play experience to revise them for mastery. Starting from an initial game, PlayCraft alternates between playing and crafting. In the playing stage, a player agent watches the game screen, uses the controls, and pursues the game objective over repeated episodes while retaining memory. After each episode, it reports its progress, outcomes, decisions, and reflections. In the crafting stage, a crafter agent uses these reports to revise the game. It estimates player abilities, predicts performance at different proficiency levels, and compares those predictions with a target learning progression. It diagnoses the game and addresses the gaps by giving players meaningful opportunities to improve through better decisions and execution. We evaluate games in fresh runs with player agents across three intelligence settings, measuring learning gain and learning slope. Compared with the initial games, PlayCraft yields larger learning gains and steeper learning slopes in most games. It also outperforms the baselines on the macro-average of each metric at every intelligence setting.

## Android developer verification will reach every install source, with a side door for power users

DevFeed: [Android developer verification will reach every install source, with a side door for power users](<https://devfeed.tech/articles/android-developer-verification-will-reach-every-install-source-with-a-side-door-for-power-users-144524.md>)

Original publisher: [Read original article](<https://www.helpnetsecurity.com/2026/11/20/android-developer-verification-power-users/>)

Author: Anamarija Pogorelec

Published: 2026-11-19T23:03:24Z

Content type: news

Language: en

Sources: [Help Net Security](<https://devfeed.tech/sources/help-net-security.md>)

Topics: [appstores](<https://devfeed.tech/topics/appstores.md>), [Google Play](<https://devfeed.tech/topics/google-play.md>), [mobile devops](<https://devfeed.tech/topics/mobile-devops.md>), [CI/CD](<https://devfeed.tech/topics/cicd.md>)

Tags: [2027](<https://devfeed.tech/tags/2027.md>), [android](<https://devfeed.tech/tags/android.md>), [android-developer-verification](<https://devfeed.tech/tags/android-developer-verification.md>), [app](<https://devfeed.tech/tags/app.md>), [app-stores](<https://devfeed.tech/tags/app-stores.md>), [apps](<https://devfeed.tech/tags/apps.md>), [country](<https://devfeed.tech/tags/country.md>), [developer](<https://devfeed.tech/tags/developer.md>), [developers](<https://devfeed.tech/tags/developers.md>), [devices](<https://devfeed.tech/tags/devices.md>), [experience](<https://devfeed.tech/tags/experience.md>), [extend](<https://devfeed.tech/tags/extend.md>), [galaxy](<https://devfeed.tech/tags/galaxy.md>), [galaxy-store](<https://devfeed.tech/tags/galaxy-store.md>), [google](<https://devfeed.tech/tags/google.md>), [identity-verification](<https://devfeed.tech/tags/identity-verification.md>), [install](<https://devfeed.tech/tags/install.md>), [mobile-devices](<https://devfeed.tech/tags/mobile-devices.md>), [mobile-security](<https://devfeed.tech/tags/mobile-security.md>), [news](<https://devfeed.tech/tags/news.md>), [power-users](<https://devfeed.tech/tags/power-users.md>), [smartphones](<https://devfeed.tech/tags/smartphones.md>), [source](<https://devfeed.tech/tags/source.md>), [users](<https://devfeed.tech/tags/users.md>), [verification](<https://devfeed.tech/tags/verification.md>)

### AI overview

Starting in January 2027, Google plans to extend developer verification worldwide across Android install sources. On certified devices, unregistered apps will lose the normal install experience; the rollout begins with Google Play, Galaxy Store, Honor, Oppo, Xiaomi, Transsion and Vivo, then expands to sideloading and open-source registries. Google says registration takes under ten minutes. Students and hobbyists can use limited distribution accounts without registration fees or government ID checks, with sharing allowed to 50 authorized devices. Teams can automate registration using two APIs. Play developers must also register apps distributed outside Play. Google describes an advanced flow for experienced users to install unregistered apps if they accept the risk, while retaining guardrails against social engineering scams. The excerpt says verification already applies at participating stores in Brazil, Indonesia, Singapore and Thailand; Google reports registered apps account for over 99% of worldwide Android installs.

### Source excerpt

Google will extend Android developer verification to every country starting in January 2027, a check that already applies at participating app stores in four countries. Developers who have not verified their identity with Google and registered their apps will see those apps lose the normal install experience on certified Android devices. The rollout opens with seven app distributors: Google Play, Galaxy Store, Honor, Oppo, Xiaomi, Transsion and Vivo. Over several months it widens to every ... More -> The post Android developer verification will reach every install source, with a side door for power users appeared first on Help Net Security.

## OpenKAGE: Open-Ended Lifelong Learning in the Open World via Knowledge-Sharing Agent-Group Evolution

DevFeed: [OpenKAGE: Open-Ended Lifelong Learning in the Open World via Knowledge-Sharing Agent-Group Evolution](<https://devfeed.tech/articles/openkage-open-ended-lifelong-learning-in-the-open-world-via-knowledge-sharing-agent-group-evolution-142775.md>)

Original publisher: [Read original article](<https://arxiv.org/abs/2611.24302>)

Author: Wenhao Yuan, Longhui Yuan, Zexi Jia, Ningzhe Shi, Jingwei Li, Shipeng Cen, Pengcheng Luo, Jie Zhou, Jinchao 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: [Multi Agent Systems](<https://devfeed.tech/topics/multi-agent-systems.md>), [Artificial Intelligence](<https://devfeed.tech/topics/ai.md>), [knowledge-management](<https://devfeed.tech/topics/knowledge-management.md>), [knowledge-engineering](<https://devfeed.tech/topics/knowledge-engineering.md>)

Tags: [advanced](<https://devfeed.tech/tags/advanced.md>), [agent](<https://devfeed.tech/tags/agent.md>), [agents](<https://devfeed.tech/tags/agents.md>), [artificial-intelligence](<https://devfeed.tech/tags/artificial-intelligence.md>), [autonomous](<https://devfeed.tech/tags/autonomous.md>), [autonomous-agents](<https://devfeed.tech/tags/autonomous-agents.md>), [capabilities](<https://devfeed.tech/tags/capabilities.md>), [cs-ai](<https://devfeed.tech/tags/cs-ai.md>), [cs-ma](<https://devfeed.tech/tags/cs-ma.md>), [evolution](<https://devfeed.tech/tags/evolution.md>), [experience](<https://devfeed.tech/tags/experience.md>), [exploration](<https://devfeed.tech/tags/exploration.md>), [framework](<https://devfeed.tech/tags/framework.md>), [graph](<https://devfeed.tech/tags/graph.md>), [groups](<https://devfeed.tech/tags/groups.md>), [improvement](<https://devfeed.tech/tags/improvement.md>), [knowledge](<https://devfeed.tech/tags/knowledge.md>), [knowledge-sharing](<https://devfeed.tech/tags/knowledge-sharing.md>), [learning](<https://devfeed.tech/tags/learning.md>), [open](<https://devfeed.tech/tags/open.md>), [sharing](<https://devfeed.tech/tags/sharing.md>), [world](<https://devfeed.tech/tags/world.md>)

### AI overview

OpenKAGE enables groups of agents to keep exploring open worlds and build on one another's experience. It separates situated memories from transferable experiences and skills, retaining conditions for when those skills apply. Textual experiences are consolidated into a shared knowledge graph, while executable skills are shared verbatim. In Minecraft, the framework reportedly sustained group exploration and supported advanced technological milestones. Its accumulated knowledge also improved survival and exploration when transferred to a newly initialized group using a different backbone model.

### Source excerpt

arXiv:2611.24302v1 Announce Type: new Abstract: Building autonomous agents that continually explore open worlds and acquire new capabilities through experience is a long-standing aspiration in artificial intelligence. Existing research often pursues lifelong learning within predefined tasks, curricula, or objectives, with self-improvement largely confined to individual agents. We introduce OpenKAGE, an agent-group evolution framework for open-ended lifelong learning. Within this framework, agents autonomously explore open worlds and continually acquire new capabilities by accumulating and reusing knowledge from their own interactions and the experience of others. OpenKAGE builds on two key mechanisms: (1) a two-layer knowledge representation, with a memory layer for situated memories and an experience layer for transferable experience and skills that retains the applicability conditions needed for reuse; and (2) a knowledge-sharing mechanism that semantically consolidates textual experience into a continuously evolving shared knowledge graph, while executable skills are shared verbatim. In Minecraft, OpenKAGE supports open-ended lifelong learning, enabling agent groups to sustain exploration and reach advanced technological milestones. Furthermore, the accumulated knowledge can be transferred to a newly initialized agent group on a different backbone model, improving its survival and exploration progress.

## F$^3$: What Makes Good Experience in World-Action Models?

DevFeed: [F$^3$: What Makes Good Experience in World-Action Models?](<https://devfeed.tech/articles/f-3-what-makes-good-experience-in-world-action-models-142824.md>)

Original publisher: [Read original article](<https://arxiv.org/abs/2611.24708>)

Author: Zihe Huang, Sirui Zhou, Haozhi Han, Ruge Zhang, Teng Ma, Tao Ma, Zheng Liu, Ce Zhang, Ting Cao, Yunxin Liu, Yunquan Zhang, Yifeng Chen, Kun Li

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

Content type: article

Language: en

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

Topics: [World models](<https://devfeed.tech/topics/world-models.md>), [AI Context Management](<https://devfeed.tech/topics/ai-context-management.md>), [Cosmos](<https://devfeed.tech/topics/cosmos.md>)

Tags: [actions](<https://devfeed.tech/tags/actions.md>), [context](<https://devfeed.tech/tags/context.md>), [cosmos](<https://devfeed.tech/tags/cosmos.md>), [cosmos-3](<https://devfeed.tech/tags/cosmos-3.md>), [cs-ai](<https://devfeed.tech/tags/cs-ai.md>), [examples](<https://devfeed.tech/tags/examples.md>), [experience](<https://devfeed.tech/tags/experience.md>), [failed](<https://devfeed.tech/tags/failed.md>), [failures](<https://devfeed.tech/tags/failures.md>), [form](<https://devfeed.tech/tags/form.md>), [fusion](<https://devfeed.tech/tags/fusion.md>), [helps](<https://devfeed.tech/tags/helps.md>), [images](<https://devfeed.tech/tags/images.md>), [library](<https://devfeed.tech/tags/library.md>), [model](<https://devfeed.tech/tags/model.md>), [models](<https://devfeed.tech/tags/models.md>), [nano](<https://devfeed.tech/tags/nano.md>), [native](<https://devfeed.tech/tags/native.md>), [policy](<https://devfeed.tech/tags/policy.md>), [raises](<https://devfeed.tech/tags/raises.md>), [world](<https://devfeed.tech/tags/world.md>)

### AI overview

F$^3$ improves how world-action models use experience by deciding what to store, when to retrieve it, and how to feed it into the policy. Form turns successful and failed rollouts into short, verified operation records; Fit trains the policy to read records and predict when they will help; Fusion adds their images, text, and actions as a prefix and action prior. After training once on 200 examples, the model uses a growing experience library and reports state-of-the-art results on MolmoSpaces and RoboLab, without training on RoboLab. Compared with native Cosmos 3 Nano, it reports success gains of 12.0 and 10.8 percentage points, 12.2% and 11.3% fewer steps to success, and 57% fewer regressions. These figures are from the supplied abstract.

### Source excerpt

arXiv:2611.24708v1 Announce Type: new Abstract: Deployed world-action models (WAMs) can keep improving only if they make good use of their own experience. Existing methods supply retrieved rollouts or demonstrations to the policy as context, yet such experience often turns previously solved tasks into failures. What, then, makes good experience for a WAM? We show that experience is valuable only in use: its value depends not on whether the source rollout succeeded but on what is stored, when it is used, and where it enters the policy. We therefore propose F$^3$, which makes these three decisions with Form, Fit, and Fusion. Form compresses successful and failed rollouts alike into short, verified operation records; Fit teaches the policy to read a record and to predict when it helps; Fusion injects the record's images, text and actions into the policy as a prefix and an action prior. Trained once on only 200 examples and deployed with a growing experience library, the model achieves state-of-the-art results on MolmoSpaces and RoboLab without any training on RoboLab: over native Cosmos 3 Nano, it raises success by 12.0 pp and 10.8 pp, reduces steps to success by 12.2% and 11.3%, and cuts regressions by 57%.

## From Brisbane to the rOpenSci Community: Ten Years of Open Research Software

DevFeed: [From Brisbane to the rOpenSci Community: Ten Years of Open Research Software](<https://devfeed.tech/articles/from-brisbane-to-the-ropensci-community-ten-years-of-open-research-software-144781.md>)

Original publisher: [Read original article](<https://ropensci.org/blog/2026/11/19/brisbane-to-community/>)

Author: rOpenSci

Published: 2026-11-19T00:00:00Z

Content type: article

Language: en

Sources: [rOpenSci](<https://devfeed.tech/sources/ropensci-open-tools-for-open-science.md>)

Topics: [Open Source](<https://devfeed.tech/topics/open-source.md>), [experience report](<https://devfeed.tech/topics/experience-report.md>), [Test-driven development](<https://devfeed.tech/topics/tdd.md>)

Tags: [2016](<https://devfeed.tech/tags/2016.md>), [about](<https://devfeed.tech/tags/about.md>), [also](<https://devfeed.tech/tags/also.md>), [australia](<https://devfeed.tech/tags/australia.md>), [but](<https://devfeed.tech/tags/but.md>), [code](<https://devfeed.tech/tags/code.md>), [community](<https://devfeed.tech/tags/community.md>), [confidence](<https://devfeed.tech/tags/confidence.md>), [connect](<https://devfeed.tech/tags/connect.md>), [contributed](<https://devfeed.tech/tags/contributed.md>), [data](<https://devfeed.tech/tags/data.md>), [developing](<https://devfeed.tech/tags/developing.md>), [doing](<https://devfeed.tech/tags/doing.md>), [event](<https://devfeed.tech/tags/event.md>), [experience](<https://devfeed.tech/tags/experience.md>), [first](<https://devfeed.tech/tags/first.md>), [open](<https://devfeed.tech/tags/open.md>), [research](<https://devfeed.tech/tags/research.md>), [ropensci](<https://devfeed.tech/tags/ropensci.md>), [software](<https://devfeed.tech/tags/software.md>)

### AI overview

The author's ten-year rOpenSci story shows how a welcoming, hands-on research software community can help newcomers contribute and grow. At the 2016 Brisbane Unconf, they improved code for downloading Australian weather grids and wrote a vignette; subsequent software peer review helped strengthen GSODR's focus, reproducibility and robustness, and taught them the importance of documentation, testing and ongoing maintenance. They later developed weather and climate packages, learned to test remote APIs with {vcr}, reviewed packages and became an editor. The author says reviewing others' code also improved their own practice. Lessons from rOpenSci informed Open Plant Pathology, which the founders wound down in 2025. They encourage contributions at many levels, from bug reports and documentation to package reviews, and argue that sustained community relationships and constructive feedback matter alongside the software. They note that rOpenSci must adapt to changing tools and growing editorial demands while keeping peer review useful and welcoming.

### Source excerpt

With rOpenSci turning 15 this year, it's hard to believe that I've been part of it for 10 of those years. My involvement started in Brisbane in 2016, at the first Australian rOpenSci Unconference. I'd known about rOpenSci and its work for some time before then. When rOpenSci was founded in 2011, I'd just moved from the US to the Philippines for work. During the five years I lived and worked there, I became familiar with its packages and the people behind them. I was interested in what rOpenSci was doing, but wasn't sure how I might get involved. I did recognise the value of being part of the R community, though, and joined the Manila R User Group. Brisbane 2016: just the beginning I moved to Australia for work in January 2016 and, as luck would have it, was living close enough to Brisbane to attend the rOpenSci Unconf in April. The event was organised by a group of researchers at QUT, including Nick Tierney, Miles McBain and Jessie Roberts. As it happened, my first R package, {GSODR}, had just been accepted onto CRAN. That was the fact I shared when introducing myself to the other attendees. I was an experienced researcher and R user, but still very much a newcomer to developing and publishing R packages. I was nervous about attending. I was new to Australia and had never been to an event quite like it. But the chance to bring Australian R users together to work on practical research problems was exciting, and I soon settled in. The format was very hands-on. We shared unfinished ideas, learned from one another and built things together. I found this a great way to learn, and it helped me gain confidence. I was one of the few attendees with experience working with geospatial data and a particular interest in weather data, which led me to connect with A/Prof Ivan Hanigan. Ivan had some R code for downloading AWAP grids, AWAP_GRIDS, and I was able to help improve it. I also contributed a vignette describing how to work with Australian weather data, which went on to inf

## Why we started Context

DevFeed: [Why we started Context](<https://devfeed.tech/articles/why-we-started-context-144648.md>)

Original publisher: [Read original article](<https://www.mintlify.com/blog/why-we-started-context>)

Author: Chloe Wu-Breshears

Published: 2026-11-19T00:00:00Z

Content type: news

Language: en

Sources: [Mintlify Blog](<https://devfeed.tech/sources/mintlify-blog.md>)

Topics: [Documentation](<https://devfeed.tech/topics/documentation.md>), [UI/UX Design](<https://devfeed.tech/topics/ui-ux-design.md>), [knowledge-management](<https://devfeed.tech/topics/knowledge-management.md>), [context](<https://devfeed.tech/topics/context.md>)

Tags: [agents](<https://devfeed.tech/tags/agents.md>), [context](<https://devfeed.tech/tags/context.md>), [devrel](<https://devfeed.tech/tags/devrel.md>), [docs](<https://devfeed.tech/tags/docs.md>), [documentation](<https://devfeed.tech/tags/documentation.md>), [editor](<https://devfeed.tech/tags/editor.md>), [event](<https://devfeed.tech/tags/event.md>), [events](<https://devfeed.tech/tags/events.md>), [experience](<https://devfeed.tech/tags/experience.md>)

### AI overview

At Mintlify's first Context event, documentation leaders discussed how to serve both human readers and agents as agents account for almost 80% of docs traffic, according to the excerpt. They identified a "Goldilocks" challenge: too little context leaves readers guessing, while too much obscures what matters, and the right amount may vary by reader. The discussion also raised open questions about whether people will continue to value polished, rendered docs alongside faster reading tools, and whether knowledge will shift toward plain Markdown. Panelists suggested that documentation formats may adapt to how each reader learns, while emphasizing that finding, reading and acting on information makes UX central to docs. The excerpt presents these as current questions rather than settled predictions.

### Source excerpt

A recap of our first evening at Letterform Archive, where technical writers, DevRel and engineering leaders talked about how people and agents will read, write and share knowledge.

## Dell XPS 14 DA14260 Powered By Intel Panther Lake Delivers Nice Linux Experience

DevFeed: [Dell XPS 14 DA14260 Powered By Intel Panther Lake Delivers Nice Linux Experience](<https://devfeed.tech/articles/dell-xps-14-da14260-powered-by-intel-panther-lake-delivers-nice-linux-experience-141231.md>)

Original publisher: [Read original article](<https://www.phoronix.com/review/dell-xps-14-panther-lake>)

Author: Michael Larabel

Published: 2026-11-18T21:37:15Z

Content type: article

Language: en

Sources: [Phoronix](<https://devfeed.tech/sources/phoronix.md>)

Topics: [dell](<https://devfeed.tech/topics/dell.md>), [Linux](<https://devfeed.tech/topics/linux.md>), [Ubuntu](<https://devfeed.tech/topics/ubuntu.md>), [Intel Core](<https://devfeed.tech/topics/intel-core.md>), [intel](<https://devfeed.tech/topics/intel.md>)

Tags: [3](<https://devfeed.tech/tags/3.md>), [arc](<https://devfeed.tech/tags/arc.md>), [box](<https://devfeed.tech/tags/box.md>), [core](<https://devfeed.tech/tags/core.md>), [dell](<https://devfeed.tech/tags/dell.md>), [dell-xps](<https://devfeed.tech/tags/dell-xps.md>), [desktop-linux](<https://devfeed.tech/tags/desktop-linux.md>), [efficiency](<https://devfeed.tech/tags/efficiency.md>), [experience](<https://devfeed.tech/tags/experience.md>), [graphics](<https://devfeed.tech/tags/graphics.md>), [intel](<https://devfeed.tech/tags/intel.md>), [intel-core](<https://devfeed.tech/tags/intel-core.md>), [laptop](<https://devfeed.tech/tags/laptop.md>), [latest](<https://devfeed.tech/tags/latest.md>), [linux](<https://devfeed.tech/tags/linux.md>), [linux-benchmarking](<https://devfeed.tech/tags/linux-benchmarking.md>), [linux-hardware-benchmarks](<https://devfeed.tech/tags/linux-hardware-benchmarks.md>), [linux-hardware-reviews](<https://devfeed.tech/tags/linux-hardware-reviews.md>), [linux-how-to](<https://devfeed.tech/tags/linux-how-to.md>), [linux-performance](<https://devfeed.tech/tags/linux-performance.md>), [linux-server-benchmarks](<https://devfeed.tech/tags/linux-server-benchmarks.md>), [lts](<https://devfeed.tech/tags/lts.md>), [nice](<https://devfeed.tech/tags/nice.md>), [open-source-graphics](<https://devfeed.tech/tags/open-source-graphics.md>), [panther-lake](<https://devfeed.tech/tags/panther-lake.md>), [performant](<https://devfeed.tech/tags/performant.md>), [phoronix](<https://devfeed.tech/tags/phoronix.md>), [phoronix-test-suite](<https://devfeed.tech/tags/phoronix-test-suite.md>), [power](<https://devfeed.tech/tags/power.md>), [ubuntu-benchmarks](<https://devfeed.tech/tags/ubuntu-benchmarks.md>), [ubuntu-hardware](<https://devfeed.tech/tags/ubuntu-hardware.md>), [xps](<https://devfeed.tech/tags/xps.md>)

### AI overview

Dell's XPS 14 DA14260 with Intel Panther Lake provides a capable Ubuntu 26.04 LTS experience out of the box: it runs the stock Linux 7.0 kernel without an OEM kernel, and the webcam, audio, haptic touchpad and other typical functions worked. The review unit's Arc B390 graphics and battery efficiency were strong points, with all-day battery life reported for content viewing or light browsing and office work. A newer, unofficial kernel could improve graphics performance and power efficiency; Ubuntu's official Linux 7.3 HWE stack was expected with a later 26.04 point release. The tested configuration had a Core Ultra X7 358H, 32 GB memory, 512 GB NVMe storage and a 2.8K display. Display and component options vary by configuration.

### Source excerpt

For those looking for a new laptop with Linux pre-loaded and have been eyeing the Intel Core Ultra Series 3 "Panther Lake" for its nice Xe3 Arc B390 graphics and overall power efficiency, Dell recently certified their XPS 14 (DA14260) laptop for Ubuntu 26.04 LTS Linux use. This latest Dell XPS 14 is well built and provides a nice and performant out-of-the-box Linux experience.

## Genie 4 is now available

DevFeed: [Genie 4 is now available](<https://devfeed.tech/articles/genie-4-is-now-available-141196.md>)

Original publisher: [Read original article](<https://blog.google/innovation-and-ai/models-and-research/google-deepmind/genie-4-now-available-project-genie/>)

Author: Pavan Kumar

Published: 2026-11-18T18:00:00Z

Content type: release

Language: en

Sources: [Gemini Models](<https://devfeed.tech/sources/gemini-models.md>), [Google DeepMind](<https://devfeed.tech/sources/google-deepmind.md>), [Google Labs](<https://devfeed.tech/sources/google-labs.md>)

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

Tags: [agent](<https://devfeed.tech/tags/agent.md>), [ai](<https://devfeed.tech/tags/ai.md>), [audio](<https://devfeed.tech/tags/audio.md>), [browser](<https://devfeed.tech/tags/browser.md>), [community](<https://devfeed.tech/tags/community.md>), [control](<https://devfeed.tech/tags/control.md>), [creative](<https://devfeed.tech/tags/creative.md>), [custom](<https://devfeed.tech/tags/custom.md>), [environment](<https://devfeed.tech/tags/environment.md>), [events](<https://devfeed.tech/tags/events.md>), [experience](<https://devfeed.tech/tags/experience.md>), [experimental](<https://devfeed.tech/tags/experimental.md>), [exploring](<https://devfeed.tech/tags/exploring.md>), [frontier](<https://devfeed.tech/tags/frontier.md>), [gallery](<https://devfeed.tech/tags/gallery.md>), [gemini](<https://devfeed.tech/tags/gemini.md>), [gemini-models](<https://devfeed.tech/tags/gemini-models.md>), [generate](<https://devfeed.tech/tags/generate.md>), [genie](<https://devfeed.tech/tags/genie.md>), [genie-4](<https://devfeed.tech/tags/genie-4.md>), [google-deepmind](<https://devfeed.tech/tags/google-deepmind.md>), [google-labs](<https://devfeed.tech/tags/google-labs.md>), [introduction](<https://devfeed.tech/tags/introduction.md>), [none](<https://devfeed.tech/tags/none.md>)

### AI overview

Project Genie gives eligible Google AI Ultra subscribers browser-based tools to generate, explore, and remix interactive worlds. Genie 4 supports real-time prompt-driven changes during exploration and native audio that adapts to the scene. Genie 3 remains an option for longer sessions, panorama-seeded worlds, or linked journeys: generations can last up to three minutes, and prompts can trigger in-world changes. A conversational agent can coordinate multimodal models to create visuals and background scores; Gemini can offer context-aware suggestions or trigger events. Users can explore curated and community worlds, remix them, and publish their creations. Google AI Pro members can join a waitlist. Generated audio and video are SynthID-watermarked and inputs and outputs are safety-screened. The models remain frontier research, with challenges including camera latency, complex physical interactions, audio precision, and long-term scene permanence.

### Source excerpt

Introduction to Genie 4

## When the World Is the Test Suite: Agentic Intelligence for the Physical World

DevFeed: [When the World Is the Test Suite: Agentic Intelligence for the Physical World](<https://devfeed.tech/articles/when-the-world-is-the-test-suite-agentic-intelligence-for-the-physical-world-139601.md>)

Original publisher: [Read original article](<https://arxiv.org/abs/2611.22291>)

Author: Kaixin Li, Meng Luo, Yanlin Li, Yuchen Tian, Xiang Liu, Xiangru Jian, Yan Yang, Zhiheng Xi, Yang Zhang, Ziyang Luo, Mingfei Cheng, Hongzhan Lin

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

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: [Physical AI](<https://devfeed.tech/topics/physical-ai.md>), [Robotics Simulation](<https://devfeed.tech/topics/robotics-simulation.md>), [AI-assisted coding](<https://devfeed.tech/topics/ai-assisted-coding.md>), [Test automation](<https://devfeed.tech/topics/test-automation.md>)

Tags: [actuator](<https://devfeed.tech/tags/actuator.md>), [agentic](<https://devfeed.tech/tags/agentic.md>), [agents](<https://devfeed.tech/tags/agents.md>), [benchmarks](<https://devfeed.tech/tags/benchmarks.md>), [code](<https://devfeed.tech/tags/code.md>), [coding](<https://devfeed.tech/tags/coding.md>), [coding-agents](<https://devfeed.tech/tags/coding-agents.md>), [continuous-integration](<https://devfeed.tech/tags/continuous-integration.md>), [controllers](<https://devfeed.tech/tags/controllers.md>), [cost](<https://devfeed.tech/tags/cost.md>), [cs-ai](<https://devfeed.tech/tags/cs-ai.md>), [cs-ro](<https://devfeed.tech/tags/cs-ro.md>), [cs-se](<https://devfeed.tech/tags/cs-se.md>), [evaluation](<https://devfeed.tech/tags/evaluation.md>), [experience](<https://devfeed.tech/tags/experience.md>), [formal](<https://devfeed.tech/tags/formal.md>), [industrial](<https://devfeed.tech/tags/industrial.md>), [intelligence](<https://devfeed.tech/tags/intelligence.md>), [language-models](<https://devfeed.tech/tags/language-models.md>), [learning](<https://devfeed.tech/tags/learning.md>), [libraries](<https://devfeed.tech/tags/libraries.md>), [test](<https://devfeed.tech/tags/test.md>), [world](<https://devfeed.tech/tags/world.md>)

### AI overview

The paper frames physical-world agentic intelligence around a central challenge: programs for robots and other physical systems lack a built-in test suite, and trustworthy evidence of their behavior is costly and may be flawed. It surveys how agents write robot policies and programs that support policy learning, how physical trials compare with traces, learned judges, simulators, and formal checks, and why safety assessment must go beyond task success. It also examines how tested programs become reusable experience through skill libraries and model weights, alongside benchmarks and open evaluation challenges. The authors argue for a physical counterpart to continuous integration: verification that is cheap to run often, reliable during search, and maintained with the experience it validates.

### Source excerpt

arXiv:2611.22291v1 Announce Type: cross Abstract: Coding agents built to repair software now write, test, and repair programs for robots, laboratories, and industrial controllers. Unlike in software engineering, no test suite comes with the code. Whether a program works must be established by running it on a robot, in a physics simulator, or through the policies it helps to train, and such evidence is costly to obtain and can itself be wrong. As these agents take on more physical tasks, how their programs are judged matters as much as how they are written. This paper examines this emerging area, agentic intelligence for the physical world, by treating the world as the program's test suite. We organize the discussion around three questions: how agents write programs for the physical world, how these programs are tested, and how agents learn from the results. We first trace how language models for robots and software engineering agents came together, ask which software testing practices carry over to physical tests, and describe each program by where it sits relative to the actuator and by what judges it. For writing, we cover programs that act as robot policies and programs that support policy learning by producing rewards, tasks, and training worlds. For testing, we examine what physical trials cost, what traces, learned judges, simulators, and formal checks can establish in their place, and what safety requires beyond task success. For learning, we discuss how tested programs are retained as experience, in skill libraries and in model weights. Finally, we turn to benchmarks and evaluation practice and outline open challenges. Writing programs is becoming cheap, while trustworthy evidence about what they do remains costly. Progress will therefore depend on a physical counterpart of continuous integration: verification cheap enough to run often, reliable enough to trust under search, and maintained alongside the experience it validates.

## Do LLM Agents Still Need to Be Told How to Self-Improve?

DevFeed: [Do LLM Agents Still Need to Be Told How to Self-Improve?](<https://devfeed.tech/articles/do-llm-agents-still-need-to-be-told-how-to-self-improve-140135.md>)

Original publisher: [Read original article](<https://arxiv.org/abs/2611.23219>)

Author: Shobhita Sundaram, Cansu Sancaktar, Phillip Isola

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

Content type: article

Language: en

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

Topics: [Machine learning](<https://devfeed.tech/topics/machine-learning.md>), [Large Language Model](<https://devfeed.tech/topics/llm.md>)

Tags: [3d](<https://devfeed.tech/tags/3d.md>), [agents](<https://devfeed.tech/tags/agents.md>), [autonomous](<https://devfeed.tech/tags/autonomous.md>), [benchmarks](<https://devfeed.tech/tags/benchmarks.md>), [compute](<https://devfeed.tech/tags/compute.md>), [cs-ai](<https://devfeed.tech/tags/cs-ai.md>), [cs-lg](<https://devfeed.tech/tags/cs-lg.md>), [experience](<https://devfeed.tech/tags/experience.md>), [gameplay](<https://devfeed.tech/tags/gameplay.md>), [learning](<https://devfeed.tech/tags/learning.md>), [llm](<https://devfeed.tech/tags/llm.md>), [llm-agents](<https://devfeed.tech/tags/llm-agents.md>), [optimization](<https://devfeed.tech/tags/optimization.md>), [self-improvement](<https://devfeed.tech/tags/self-improvement.md>), [strategies](<https://devfeed.tech/tags/strategies.md>), [tools](<https://devfeed.tech/tags/tools.md>)

### AI overview

The study finds that sufficiently capable frozen LLM agents can improve at gameplay when given a compute budget, experience-collecting tools, editable text documents, and a general goal, without an imposed learning loop or access to real-world data. The agents construct their own context-based learning strategies, which the authors say align with frameworks for self-regulated learning and self-generated curricula. The learned text artifacts also transfer to external benchmarks, with reported gains across game horizons, mechanics, and 3D navigation. The evidence here is an abstract, so it does not specify model details, the size of the gains, or how broadly the results generalize.

### Source excerpt

arXiv:2611.23219v1 Announce Type: cross Abstract: LLM agents increasingly drive their own learning. Self-improvement methods delegate various components of model optimization to the models themselves. As models become more capable, this motivates a basic question: do these increasingly capable models still need us to specify how they should self-improve? In this paper we study an autonomous paradigm in which, given basic primitives, resources, and a general goal for self-improvement, agents construct their own self-learning algorithm for context-based optimization. Using the setting of gameplay as a testbed, a frozen LLM agent is given a compute budget and tools for collecting experience and updating text documents, and allowed to freely experiment without an externally imposed loop or access to real-world data. We find that sufficiently capable models do self-improve under these conditions, demonstrating strategies that align with classic cognitive frameworks for self-regulated learning and constructs from machine learning literature on self-generated curricula. Furthermore, we study the extent to which the learned text artifacts transfer to external benchmarks, observing improvement across different game horizons, mechanics, and 3D navigation.

## What Does an Agent Need? Bundle-Aware Agent Resource Composition

DevFeed: [What Does an Agent Need? Bundle-Aware Agent Resource Composition](<https://devfeed.tech/articles/what-does-an-agent-need-bundle-aware-agent-resource-composition-139949.md>)

Original publisher: [Read original article](<https://arxiv.org/abs/2611.22030>)

Author: Yunxiao Shi (Australian National University), Tingwei Chen (University of Technology Sydney), Dannong Xu (Australian National University), Ahad N. Zehmakan (Australian National University), Jialie She

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

Content type: article

Language: en

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

Topics: [AI Agent](<https://devfeed.tech/topics/ai-agent.md>), [Model Routing](<https://devfeed.tech/topics/model-routing.md>), [Retrieval Augmented Generation (RAG)](<https://devfeed.tech/topics/retrieval-augmented-generation-rag.md>), [Large Language Model](<https://devfeed.tech/topics/llm.md>), [Fine-tuning](<https://devfeed.tech/topics/fine-tuning.md>)

Tags: [agent](<https://devfeed.tech/tags/agent.md>), [agents](<https://devfeed.tech/tags/agents.md>), [beam](<https://devfeed.tech/tags/beam.md>), [candidate](<https://devfeed.tech/tags/candidate.md>), [complete](<https://devfeed.tech/tags/complete.md>), [composition](<https://devfeed.tech/tags/composition.md>), [configuration](<https://devfeed.tech/tags/configuration.md>), [cs-ai](<https://devfeed.tech/tags/cs-ai.md>), [execution](<https://devfeed.tech/tags/execution.md>), [experience](<https://devfeed.tech/tags/experience.md>), [feedback](<https://devfeed.tech/tags/feedback.md>), [fine](<https://devfeed.tech/tags/fine.md>), [generation](<https://devfeed.tech/tags/generation.md>), [generator](<https://devfeed.tech/tags/generator.md>), [generators](<https://devfeed.tech/tags/generators.md>), [grounded](<https://devfeed.tech/tags/grounded.md>), [language](<https://devfeed.tech/tags/language.md>), [language-model](<https://devfeed.tech/tags/language-model.md>), [large](<https://devfeed.tech/tags/large.md>), [large-language-model](<https://devfeed.tech/tags/large-language-model.md>), [resource](<https://devfeed.tech/tags/resource.md>)

### AI overview

AgentComposition jointly selects a backbone model and tool-skill bundle for a task, addressing the limits of generators tied to a training catalog and methods that score resources independently. Its retrieval-grounded generator combines candidate resources and past configurations through constrained beam search, while a bundle-aware critic ranks complete configurations using an LLM teacher and successful historical configurations as execution proxies; it can also adapt to execution feedback. On AgentSelect queries, the method recovered the full reference configuration within ten candidates 67.73% of the time, nearly twice the strongest baseline. The critic raised top-agent execution scores from 7.58 to 8.37 out of 10, and feedback raised them to 8.78. Four target-domain queries adapted the generator to SkillsBench, where the method led baselines. Online recommendations also outperformed the strongest baselines under an LLM judge and native verifiers.

### Source excerpt

arXiv:2611.22030v1 Announce Type: new Abstract: An agent resource's value depends on its backbone and companion resources and is known only after execution. Token-based generators remain tied to their training catalog, while independent resource scoring misses complementarities and reused bundles target other tasks. We formulate Task-Conditioned Agent Resource Composition (AgentComposition): jointly selecting a backbone model and a tool-skill bundle to maximize end-to-end execution utility. Our Retrieval-Grounded Bundle-Aware Agent Generation retrieves candidate resources and past configurations of similar tasks. A fine-tuned generator reads resource descriptions and recombines candidates through constrained beam search, allowing new resources without retraining. A bundle-aware critic ranks complete configurations using a large language model (LLM) teacher and historically successful configurations as execution proxies, and adapts to execution feedback when available. We evaluate reference recovery, offline execution with judged and native rewards, and online recommendation with real execution, building AgentSelect-Exec with 5,060 executed configurations. Our method recovers the complete reference configuration within ten candidates for 67.73% of AgentSelect queries, nearly twice the strongest baseline, and leads all baselines on SkillsBench. The proxy-trained critic raises the top-ranked agent's execution score from 7.58 to 8.37 out of 10; execution feedback increases it to 8.78, above the strongest baselines. Four target-domain queries suffice to adapt the generator to SkillsBench. Agents recommended online outperform the strongest baselines under both an LLM judge and native verifiers. These results support learning and transferring joint resource composition from task experience.

## Aeotec joins Works with Home Assistant

DevFeed: [Aeotec joins Works with Home Assistant](<https://devfeed.tech/articles/aeotec-joins-works-with-home-assistant-140870.md>)

Original publisher: [Read original article](<https://www.home-assistant.io/blog/2026/11/18/aeotec-joins-works-with-home-assistant/>)

Author: Jacqueline Ajayi

Published: 2026-11-18T00:00:01Z

Content type: news

Language: en

Sources: [Home Assistant](<https://devfeed.tech/sources/home-assistant.md>)

Topics: [Home Assistant](<https://devfeed.tech/topics/home-assistant.md>), [smartdevices](<https://devfeed.tech/topics/smartdevices.md>), [Matter](<https://devfeed.tech/topics/matter.md>)

Tags: [2006](<https://devfeed.tech/tags/2006.md>), [2016](<https://devfeed.tech/tags/2016.md>), [aeon](<https://devfeed.tech/tags/aeon.md>), [certification](<https://devfeed.tech/tags/certification.md>), [community](<https://devfeed.tech/tags/community.md>), [connectivity](<https://devfeed.tech/tags/connectivity.md>), [devices](<https://devfeed.tech/tags/devices.md>), [ecosystem](<https://devfeed.tech/tags/ecosystem.md>), [experience](<https://devfeed.tech/tags/experience.md>), [gear](<https://devfeed.tech/tags/gear.md>), [hardware](<https://devfeed.tech/tags/hardware.md>), [hub](<https://devfeed.tech/tags/hub.md>), [innovation](<https://devfeed.tech/tags/innovation.md>), [integration](<https://devfeed.tech/tags/integration.md>), [internet-of-things](<https://devfeed.tech/tags/internet-of-things.md>), [internet-of-things-iot](<https://devfeed.tech/tags/internet-of-things-iot.md>), [iot](<https://devfeed.tech/tags/iot.md>), [join](<https://devfeed.tech/tags/join.md>), [labs](<https://devfeed.tech/tags/labs.md>), [works-with-home-assistant](<https://devfeed.tech/tags/works-with-home-assistant.md>)

### AI overview

Aeotec has joined Works with Home Assistant with nine certified devices, including five Z-Wave Long Range products. Long Range uses a star network with direct adapter connections, offering up to 1 km line-of-sight range and support for up to 4,000 devices, compared with about 300 m and 232 devices for standard Z-Wave mesh. Using Long Range requires a compatible adapter such as Home Assistant Connect ZWA-2; older adapters can still use the devices as standard Z-Wave mesh devices. The devices run locally through Home Assistant's Z-Wave integration and Z-Wave JS. Setup supports QR-code SmartStart pairing and built-in S2 encryption. The certification program tests for local control and privacy, and Aeotec has committed to long-term support and firmware updates. The listed lineup covers sensors, a smoke detector, switches, and a siren; the article says one Zigbee device was tested separately with the ZBT-2 and Zigbee Home Automation integration.

### Source excerpt

Get ready for gear that goes the distance, because this month Aeotec join Works with Home Assistant 🎉 Building Z-Wave and smart home tech since 2006, our newest partners bring 9 certified devices to the program. From sensors and sirens to switches and a smoke detector upgrade - not to mention new Z-Wave Long Range devices from the 8 series - there's plenty of exciting ground to cover 📡 Howdy, partner! Aeotec are no strangers to the Home Assistant community, and for good reason. Founded in Silicon Valley in 2006 as Aeon Labs, they've spent two decades building Z-Wave and other smart home hardware. In 2016, Aeon Labs rebranded as Aeotec. Later, Aeotec merged with Z-Wave Europe to create an Internet of Things (IoT) powerhouse with offices and expertise across three continents, expanding and accelerating innovation even further. While you might know them as the official hardware partner behind the Samsung SmartThings hub, Aeotec's devices travel well beyond any one platform. Instead of locking you into a closed ecosystem, they build across open connectivity standards like Z-Wave, Zigbee, and Matter, so you can choose how best to set up your smart home - something we always love to see in our partners! Aeotec's latest devices are all about helping your smart home stretch further with Z-Wave Long Range (more on that in a minute). But plenty of our Home Assistant community members have been running Aeotec devices through the Z-Wave integration for years, so this certification is less of an introduction and more of a long-overdue handshake. We're thrilled to make it official 🎉 "At Aeotec, we have always built our products together with the community, not just for it. Smart home enthusiasts have shaped our devices from the very beginning, and we feel a deep responsibility to give them a flawless experience. Home Assistant is one of the most important pillars of that community, so being part of the Works with Home Assistant program means a great deal to us. It is our commitme

## Putting a spotlight on AI, society, and humanity

DevFeed: [Putting a spotlight on AI, society, and humanity](<https://devfeed.tech/articles/putting-a-spotlight-on-ai-society-and-humanity-139673.md>)

Original publisher: [Read original article](<https://news.mit.edu/2026/putting-spotlight-ai-society-humanity-1117>)

Author: Danna Lorch | School of Humanities, Arts, and Social Sciences

Published: 2026-11-17T20:40:00Z

Content type: news

Language: en

Sources: [MIT AI News](<https://devfeed.tech/sources/mit-ai-news.md>)

Topics: [Artificial Intelligence](<https://devfeed.tech/topics/ai.md>), [Machine learning](<https://devfeed.tech/topics/machine-learning.md>), [User Experience](<https://devfeed.tech/topics/user-experience.md>), [human computer interaction](<https://devfeed.tech/topics/human-computer-interaction.md>)

Tags: [adam-berinsky](<https://devfeed.tech/tags/adam-berinsky.md>), [agustin-rayo](<https://devfeed.tech/tags/agustin-rayo.md>), [ai](<https://devfeed.tech/tags/ai.md>), [ai-and-human-cognition](<https://devfeed.tech/tags/ai-and-human-cognition.md>), [ai-and-society](<https://devfeed.tech/tags/ai-and-society.md>), [ai-and-writing](<https://devfeed.tech/tags/ai-and-writing.md>), [ai-in-education](<https://devfeed.tech/tags/ai-in-education.md>), [ai-in-music](<https://devfeed.tech/tags/ai-in-music.md>), [ai-in-politics](<https://devfeed.tech/tags/ai-in-politics.md>), [anantha-chandrakasan](<https://devfeed.tech/tags/anantha-chandrakasan.md>), [anna-huang](<https://devfeed.tech/tags/anna-huang.md>), [anthropology](<https://devfeed.tech/tags/anthropology.md>), [artificial-intelligence](<https://devfeed.tech/tags/artificial-intelligence.md>), [arts](<https://devfeed.tech/tags/arts.md>), [arvind-satyanarayan](<https://devfeed.tech/tags/arvind-satyanarayan.md>), [brain-and-cognitive-sciences](<https://devfeed.tech/tags/brain-and-cognitive-sciences.md>), [chara-podimata](<https://devfeed.tech/tags/chara-podimata.md>), [charles-stewart-iii](<https://devfeed.tech/tags/charles-stewart-iii.md>), [collaboration](<https://devfeed.tech/tags/collaboration.md>), [comparative-media-studies-writing](<https://devfeed.tech/tags/comparative-media-studies-writing.md>), [dwai-banerjee](<https://devfeed.tech/tags/dwai-banerjee.md>), [education](<https://devfeed.tech/tags/education.md>), [education-teaching-academics](<https://devfeed.tech/tags/education-teaching-academics.md>), [electrical-engineering-and-computer-science-eecs](<https://devfeed.tech/tags/electrical-engineering-and-computer-science-eecs.md>), [eric-klopfer](<https://devfeed.tech/tags/eric-klopfer.md>), [ethics](<https://devfeed.tech/tags/ethics.md>), [experience](<https://devfeed.tech/tags/experience.md>), [funding](<https://devfeed.tech/tags/funding.md>), [graham-jones](<https://devfeed.tech/tags/graham-jones.md>), [higher-education](<https://devfeed.tech/tags/higher-education.md>), [human-computer-interaction](<https://devfeed.tech/tags/human-computer-interaction.md>), [human-insight-collaborative](<https://devfeed.tech/tags/human-insight-collaborative.md>), [humanities](<https://devfeed.tech/tags/humanities.md>), [humanity](<https://devfeed.tech/tags/humanity.md>), [joe-paradiso](<https://devfeed.tech/tags/joe-paradiso.md>), [mariel-garcia-montes](<https://devfeed.tech/tags/mariel-garcia-montes.md>), [media-lab](<https://devfeed.tech/tags/media-lab.md>), [michael-trice](<https://devfeed.tech/tags/michael-trice.md>), [midterm-elections](<https://devfeed.tech/tags/midterm-elections.md>), [mit](<https://devfeed.tech/tags/mit.md>), [mit-class-6-s061-21a-s02-humane-user-experience-design](<https://devfeed.tech/tags/mit-class-6-s061-21a-s02-humane-user-experience-design.md>), [mit-generative-ai-impact-consortium-mgaic](<https://devfeed.tech/tags/mit-generative-ai-impact-consortium-mgaic.md>), [mit-human-insight-collaborative-mithic](<https://devfeed.tech/tags/mit-human-insight-collaborative-mithic.md>), [mit-schwarzman-college-of-computing](<https://devfeed.tech/tags/mit-schwarzman-college-of-computing.md>), [mit-shass](<https://devfeed.tech/tags/mit-shass.md>), [mit-sloan-school-of-management](<https://devfeed.tech/tags/mit-sloan-school-of-management.md>), [music-and-theater-arts](<https://devfeed.tech/tags/music-and-theater-arts.md>), [neuroscience](<https://devfeed.tech/tags/neuroscience.md>), [pawan-sinha](<https://devfeed.tech/tags/pawan-sinha.md>), [political-science](<https://devfeed.tech/tags/political-science.md>), [program-in-sts](<https://devfeed.tech/tags/program-in-sts.md>), [rhetoric](<https://devfeed.tech/tags/rhetoric.md>), [sally-kornbluth](<https://devfeed.tech/tags/sally-kornbluth.md>), [school-of-engineering](<https://devfeed.tech/tags/school-of-engineering.md>), [school-of-humanities-arts-and-social-sciences](<https://devfeed.tech/tags/school-of-humanities-arts-and-social-sciences.md>), [school-of-science](<https://devfeed.tech/tags/school-of-science.md>), [social-and-ethical-responsibilities-of-computing-serc](<https://devfeed.tech/tags/social-and-ethical-responsibilities-of-computing-serc.md>), [social-sciences](<https://devfeed.tech/tags/social-sciences.md>), [society](<https://devfeed.tech/tags/society.md>), [special-events-and-guest-speakers](<https://devfeed.tech/tags/special-events-and-guest-speakers.md>), [technology-and-society](<https://devfeed.tech/tags/technology-and-society.md>), [writing](<https://devfeed.tech/tags/writing.md>)

### AI overview

MIT's Human Insight Collaborative (MITHIC) is funding interdisciplinary work to examine how AI affects learning, cognition, democracy, creative practice, and social relationships. An MIT report found students attending fewer office hours and spending less time on problem sets and collaborative work, raising concerns about AI eroding the "creative friction" of education. Among the projects described, researchers found that Alaska voters using Anthropic's Claude for election research may have received incomplete candidate information; another team is studying whether generative AI supports cognition or shifts cognitive work away from people. Musicians and researchers are testing how AI improvisation changes authorship and ensemble dynamics. MITHIC also supports a course on humane user experience design, a renewed computing and society concentration, and faculty workshops and resources on generative AI and writing. Launched in 2024, the initiative has supported 67 projects across 31 MIT units with $5.5 million; roughly one third focus on AI, society, and humanity.

### Source excerpt

Supported by the MIT Human Insight Collaborative, scholars are exploring how AI is impacting the human experience and reshaping higher education.

## UX projects for beginners: A step-by-step guide to designing on Penpot

DevFeed: [UX projects for beginners: A step-by-step guide to designing on Penpot](<https://devfeed.tech/articles/ux-projects-for-beginners-a-step-by-step-guide-to-designing-on-penpot-138144.md>)

Original publisher: [Read original article](<https://penpot.app/blog/ux-projects-for-beginners/>)

Author: Miryam González Duque

Published: 2026-11-17T17:01:00Z

Content type: tutorial

Language: en

Sources: [Penpot Blog](<https://devfeed.tech/sources/penpot-blog.md>)

Topics: [UI/UX Design](<https://devfeed.tech/topics/ui-ux-design.md>), [human centric design](<https://devfeed.tech/topics/human-centric-design.md>), [User interface design](<https://devfeed.tech/topics/ui-design.md>), [mobile-ui-design](<https://devfeed.tech/topics/mobile-ui-design.md>), [Learning](<https://devfeed.tech/topics/learning.md>)

Tags: [beginner](<https://devfeed.tech/tags/beginner.md>), [beginners](<https://devfeed.tech/tags/beginners.md>), [clickable](<https://devfeed.tech/tags/clickable.md>), [design](<https://devfeed.tech/tags/design.md>), [designing](<https://devfeed.tech/tags/designing.md>), [experience](<https://devfeed.tech/tags/experience.md>), [first](<https://devfeed.tech/tags/first.md>), [follow](<https://devfeed.tech/tags/follow.md>), [friendly](<https://devfeed.tech/tags/friendly.md>), [guide](<https://devfeed.tech/tags/guide.md>), [learn](<https://devfeed.tech/tags/learn.md>), [penpot](<https://devfeed.tech/tags/penpot.md>), [project](<https://devfeed.tech/tags/project.md>), [projects](<https://devfeed.tech/tags/projects.md>), [prototype](<https://devfeed.tech/tags/prototype.md>), [research](<https://devfeed.tech/tags/research.md>), [user](<https://devfeed.tech/tags/user.md>), [ux](<https://devfeed.tech/tags/ux.md>), [ux-design](<https://devfeed.tech/tags/ux-design.md>), [with](<https://devfeed.tech/tags/with.md>)

### AI overview

The excerpt presents a six-step UX project using a plant-watering app: research users, define needs with a persona and requirements, map the journey, wireframe screens, build a clickable prototype, then test and revise from feedback. It recommends starting with interviews, keeping early layouts grayscale, and using a concrete test task to observe where users get stuck. Penpot's Flex Layout helps components resize when text changes, while its Prototype tab connects screens with click triggers and transitions; comments and annotations capture implementation notes. The excerpt also highlights collaboration, developer inspection, open web standards, templates and export options. It begins to detail wireframing in step 3 and prototype creation in step 5, so the supplied text does not fully spell out all six steps.

### Source excerpt

Learn how to start your first UX design project with Penpot. Follow six beginner-friendly steps from user research to a clickable prototype, no experience required.

## REAPER Digital Audio Workstation Sees Native Wayland Port

DevFeed: [REAPER Digital Audio Workstation Sees Native Wayland Port](<https://devfeed.tech/articles/reaper-digital-audio-workstation-sees-native-wayland-port-136898.md>)

Original publisher: [Read original article](<https://www.phoronix.com/news/REAPER-DAW-Wayland>)

Author: Michael Larabel

Published: 2026-11-17T10:42:54Z

Content type: news

Language: en

Sources: [Phoronix](<https://devfeed.tech/sources/phoronix.md>)

Topics: [Wayland](<https://devfeed.tech/topics/wayland.md>)

Tags: [audio](<https://devfeed.tech/tags/audio.md>), [c-plus-plus](<https://devfeed.tech/tags/c-plus-plus.md>), [code](<https://devfeed.tech/tags/code.md>), [daw](<https://devfeed.tech/tags/daw.md>), [desktop-linux](<https://devfeed.tech/tags/desktop-linux.md>), [digital](<https://devfeed.tech/tags/digital.md>), [experience](<https://devfeed.tech/tags/experience.md>), [linux](<https://devfeed.tech/tags/linux.md>), [linux-benchmarking](<https://devfeed.tech/tags/linux-benchmarking.md>), [linux-hardware-benchmarks](<https://devfeed.tech/tags/linux-hardware-benchmarks.md>), [linux-hardware-reviews](<https://devfeed.tech/tags/linux-hardware-reviews.md>), [linux-how-to](<https://devfeed.tech/tags/linux-how-to.md>), [linux-performance](<https://devfeed.tech/tags/linux-performance.md>), [linux-server-benchmarks](<https://devfeed.tech/tags/linux-server-benchmarks.md>), [native](<https://devfeed.tech/tags/native.md>), [open-source-graphics](<https://devfeed.tech/tags/open-source-graphics.md>), [pending](<https://devfeed.tech/tags/pending.md>), [phoronix](<https://devfeed.tech/tags/phoronix.md>), [phoronix-test-suite](<https://devfeed.tech/tags/phoronix-test-suite.md>), [port](<https://devfeed.tech/tags/port.md>), [reaper](<https://devfeed.tech/tags/reaper.md>), [software](<https://devfeed.tech/tags/software.md>), [ubuntu-benchmarks](<https://devfeed.tech/tags/ubuntu-benchmarks.md>), [ubuntu-hardware](<https://devfeed.tech/tags/ubuntu-hardware.md>), [wayland](<https://devfeed.tech/tags/wayland.md>), [workstation](<https://devfeed.tech/tags/workstation.md>), [x11](<https://devfeed.tech/tags/x11.md>)

### AI overview

REAPER has a pending merge request for native Wayland support. The port, by Jason Donenfeld, includes work on the WDL C++ library used by REAPER and fixes needed for the application. The excerpt says the Wayland version retains the ability to load REAPER's X11 plug-ins.

### Source excerpt

The REAPER digital audio workstation software, which was ported to Linux over the past decade and targeting X11, now has pending code for a native Wayland experience...

## Experience: ENTRAP - An IaC Enterprise Cyber Range for Reproducible Security Assessment and Dataset Generation

DevFeed: [Experience: ENTRAP - An IaC Enterprise Cyber Range for Reproducible Security Assessment and Dataset Generation](<https://devfeed.tech/articles/experience-entrap-an-iac-enterprise-cyber-range-for-reproducible-security-assessment-and-dataset-generation-139307.md>)

Original publisher: [Read original article](<https://arxiv.org/abs/2611.20594>)

Author: Rathan Appana, Aditya Kulkarni, Bheemarjuna Reddy Tamma

Published: 2026-11-17T05:00:00Z

Content type: article

Language: en

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

Topics: [Security](<https://devfeed.tech/topics/security.md>), [threat detection](<https://devfeed.tech/topics/threat-detection.md>), [Threat monitoring](<https://devfeed.tech/topics/threat-monitoring.md>), [cyber resiliency](<https://devfeed.tech/topics/cyber-resiliency.md>)

Tags: [ansible](<https://devfeed.tech/tags/ansible.md>), [assessment](<https://devfeed.tech/tags/assessment.md>), [attack](<https://devfeed.tech/tags/attack.md>), [attacks](<https://devfeed.tech/tags/attacks.md>), [automation](<https://devfeed.tech/tags/automation.md>), [backup](<https://devfeed.tech/tags/backup.md>), [backups](<https://devfeed.tech/tags/backups.md>), [breach](<https://devfeed.tech/tags/breach.md>), [campaign](<https://devfeed.tech/tags/campaign.md>), [classifier](<https://devfeed.tech/tags/classifier.md>), [clients](<https://devfeed.tech/tags/clients.md>), [cs-cr](<https://devfeed.tech/tags/cs-cr.md>), [cyber](<https://devfeed.tech/tags/cyber.md>), [dataset](<https://devfeed.tech/tags/dataset.md>), [enterprise](<https://devfeed.tech/tags/enterprise.md>), [experience](<https://devfeed.tech/tags/experience.md>), [generation](<https://devfeed.tech/tags/generation.md>), [iac](<https://devfeed.tech/tags/iac.md>), [range](<https://devfeed.tech/tags/range.md>), [security](<https://devfeed.tech/tags/security.md>), [security-assessment](<https://devfeed.tech/tags/security-assessment.md>)

### AI overview

ENTRAP is an Infrastructure-as-Code cyber range for repeatable post-compromise security experiments. It uses Ansible, KVM/Libvirt, and Cloud-init to provision and reset a segmented enterprise network with services including a gateway, DMZ, DNS, DHCP, backup and web servers, and client machines. The authors run three scenarios from clean states: ransomware, Layer-2 DHCP starvation with rogue DHCP, and perimeter web exploitation. ENTRAP also generates labeled attack datasets; the authors use collected data to train and evaluate a Random Forest for ransomware detection. The abstract says the codebase, logs, and datasets are publicly released, but provides no classifier results or detailed deployment requirements.

### Source excerpt

arXiv:2611.20594v1 Announce Type: new Abstract: An enterprise breach causes most of its damage after the perimeter is crossed. Once an attacker gains a foothold, they can move laterally to other hosts, take over the trusted services of the network, and encrypt the backups needed for recovery. Defenders, on the other hand, need to study this post-compromise stage, but they cannot rehearse it on a production enterprise network. Building such an enterprise network manually is time-consuming, and resetting it to its clean state between runs is still harder. As a result, security experiments are difficult to perform, compare and reproduce. We develop Enterprise Network Testbed for Reproducible Attack and Defense Practices (ENTRAP), an Infrastructure-as-Code (IaC) cyber range that defines a fully segmented enterprise as version-controlled automation. ENTRAP uses Ansible, KVM/Libvirt, and Cloud-init to provision the environment and reset it to a clean state whenever necessary. The enterprise network includes several entities such as a gateway, an internal Demilitarized Zone (DMZ), DNS, DHCP, a backup server, a web server, and clients. On this network, a user can learn to assemble the entities, launch attacks, and write rules or defensive measures that harden its security. We launch three attacks: a Ransomware-as-a-Service (RaaS)-backed ransomware campaign, a Layer-2 DHCP starvation and rogue DHCP attack, and perimeter web exploitation. Each attack runs from a clean state, and ENTRAP acts as a reproducible generator of labeled attack datasets. Using the collected dataset, we train a Random Forest classifier for ransomware detection and present its evaluation results. We publicly release the codebase, collected logs, and datasets to facilitate further research.

## Experience Rot: Experienced Agents Fail Where Novices Succeed

DevFeed: [Experience Rot: Experienced Agents Fail Where Novices Succeed](<https://devfeed.tech/articles/experience-rot-experienced-agents-fail-where-novices-succeed-137631.md>)

Original publisher: [Read original article](<https://arxiv.org/abs/2611.20922>)

Author: Zhen Xu, Hang Yin, Zitao Zhang, Kevin Zhao, Xuefeng Liu

Published: 2026-11-17T05:00:00Z

Content type: article

Language: en

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

Topics: [AI Context Management](<https://devfeed.tech/topics/ai-context-management.md>), [Tool](<https://devfeed.tech/topics/tool.md>)

Tags: [abstract](<https://devfeed.tech/tags/abstract.md>), [after](<https://devfeed.tech/tags/after.md>), [agent](<https://devfeed.tech/tags/agent.md>), [agents](<https://devfeed.tech/tags/agents.md>), [all](<https://devfeed.tech/tags/all.md>), [arxiv](<https://devfeed.tech/tags/arxiv.md>), [call](<https://devfeed.tech/tags/call.md>), [change](<https://devfeed.tech/tags/change.md>), [context](<https://devfeed.tech/tags/context.md>), [copy](<https://devfeed.tech/tags/copy.md>), [cs-ai](<https://devfeed.tech/tags/cs-ai.md>), [cs-cl](<https://devfeed.tech/tags/cs-cl.md>), [cs-lg](<https://devfeed.tech/tags/cs-lg.md>), [definition](<https://devfeed.tech/tags/definition.md>), [documentation](<https://devfeed.tech/tags/documentation.md>), [experience](<https://devfeed.tech/tags/experience.md>), [fail](<https://devfeed.tech/tags/fail.md>), [failure](<https://devfeed.tech/tags/failure.md>), [fall](<https://devfeed.tech/tags/fall.md>), [files](<https://devfeed.tech/tags/files.md>), [form](<https://devfeed.tech/tags/form.md>), [helps](<https://devfeed.tech/tags/helps.md>), [succeed](<https://devfeed.tech/tags/succeed.md>)

### AI overview

The paper reports "Experience Rot": after MCP tool interfaces change silently, agents with prior task experience fail on most tasks while new agents mostly succeed. Experienced agents tend to repeat earlier calls instead of checking the current interface, and may report success despite failing. The pattern appears across the tested forms of memory and worsens with more experience. Experience helps when documentation is incomplete, but after a change it can leave agents worse off than having no experience. Announcing an update barely helps, and directing agents to documentation fails when it omits the needed information.

### Source excerpt

arXiv:2611.20922v1 Announce Type: cross Abstract: Agents are increasingly built to learn from experience: they keep conversation histories, memory files, retrieved notes and skills, and reuse what worked before. We show that this experience turns against them when their tools change silently. We test a range of models with tools defined as in the Model Context Protocol (MCP). After a silent change, experienced agents fail on most tasks, while agents new to the task mostly succeed. We call this failure Experience Rot. Experienced agents copy their earlier call instead of checking the current interface, even when the tool definition shows the change. Then they report success. When the documentation is incomplete, experience is what makes agents succeed. After a change, the same experience leaves them worse off than no experience at all. The failure appears in every form of memory we test and grows with the amount of experience. Simple remedies fall short: announcing the update barely helps, and pointing the agent to the documentation fails whenever the documentation does not spell out the answer.

## What Makes a Good Move? Learning to Act from Experience in Visual Worlds

DevFeed: [What Makes a Good Move? Learning to Act from Experience in Visual Worlds](<https://devfeed.tech/articles/what-makes-a-good-move-learning-to-act-from-experience-in-visual-worlds-137650.md>)

Original publisher: [Read original article](<https://arxiv.org/abs/2611.21065>)

Author: Cehao Yang, Ziyi He, Jing Yi, Jian Guo, Hui Xiong

Published: 2026-11-17T05:00:00Z

Content type: article

Language: en

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

Topics: [Reinforcement learning](<https://devfeed.tech/topics/reinforcement-learning.md>), [Machine learning](<https://devfeed.tech/topics/machine-learning.md>), [Artificial Intelligence](<https://devfeed.tech/topics/ai.md>)

Tags: [actions](<https://devfeed.tech/tags/actions.md>), [cs-ai](<https://devfeed.tech/tags/cs-ai.md>), [cs-cv](<https://devfeed.tech/tags/cs-cv.md>), [dataset](<https://devfeed.tech/tags/dataset.md>), [decision](<https://devfeed.tech/tags/decision.md>), [environment](<https://devfeed.tech/tags/environment.md>), [experience](<https://devfeed.tech/tags/experience.md>), [exploration](<https://devfeed.tech/tags/exploration.md>), [framework](<https://devfeed.tech/tags/framework.md>), [game](<https://devfeed.tech/tags/game.md>), [guidance](<https://devfeed.tech/tags/guidance.md>), [interaction](<https://devfeed.tech/tags/interaction.md>), [learning](<https://devfeed.tech/tags/learning.md>), [policy](<https://devfeed.tech/tags/policy.md>), [progress](<https://devfeed.tech/tags/progress.md>), [reinforcement-learning](<https://devfeed.tech/tags/reinforcement-learning.md>), [results](<https://devfeed.tech/tags/results.md>), [worlds](<https://devfeed.tech/tags/worlds.md>)

### AI overview

LightAct converts visual interaction experience into action-level rewards for reinforcement learning. It summarizes interaction segments into experience entries and a game-specific skill that guides collection toward broader state coverage. At each decision root, it compares candidate actions through short rollouts using a shared frozen policy, averages environment-reported score gains across continuations, and normalizes them into relative rewards. Training uses 1,024 decision roots from one source game. Across 170 GameWorld tasks in 34 games, it reports progress gains of 23.6, 29.2, and 13.0 percentage points at model scales of 2B, 4B, and 9B. The abstract says skill guidance broadens root-state coverage and gains persist across three policy optimizers; it provides no further methodological or evaluation details.

### Source excerpt

arXiv:2611.21065v1 Announce Type: cross Abstract: Learning visual policies requires turning interaction experience into guidance for individual decisions. Trajectory outcomes obscure action quality, while unguided exploration can revisit familiar situations. We present \textsc{LightAct}, an experience-learning framework that connects skill-guided exploration with policy internalization. Interaction segments are summarized into experience entries and consolidated into a game-specific skill that guides root collection toward broader state coverage. At each root, candidate actions are compared through short rollouts under a shared frozen policy. Environment-reported score gains are averaged across continuations and normalized into relative action rewards, producing reusable supervision for reinforcement learning. Policy training uses a compact dataset of just 1,024 decision roots collected in a single source game and paired with their action rewards. Across 170 GameWorld tasks spanning 34 games, \textsc{LightAct} improves overall progress by 23.6, 29.2, and 13.0 percentage points at the 2B, 4B, and 9B scales, respectively, with our strongest model competitive with reported proprietary-model results. Ablations show broader root-state coverage with skill guidance and consistent gains across three policy optimizers. Case studies show how rollouts distinguish actions with identical immediate gains.

## Feedback-Aware Residual Multi-Agent Reinforcement Learning for Timely Communication in Cognitive Radio Networks

DevFeed: [Feedback-Aware Residual Multi-Agent Reinforcement Learning for Timely Communication in Cognitive Radio Networks](<https://devfeed.tech/articles/feedback-aware-residual-multi-agent-reinforcement-learning-for-timely-communication-in-cognitive-radio-networks-137024.md>)

Original publisher: [Read original article](<https://arxiv.org/abs/2611.19229>)

Author: Baolin Chong, Xiuqi Chen, Jiandong Guo, Fuhui Zhou, Qihui Wu

Published: 2026-11-17T05:00:00Z

Content type: article

Language: en

Sources: [arXiv: Networking](<https://devfeed.tech/sources/arxiv-networking.md>)

Topics: [Reinforcement learning](<https://devfeed.tech/topics/reinforcement-learning.md>), [Multi Agent Systems](<https://devfeed.tech/topics/multi-agent-systems.md>)

Tags: [channel](<https://devfeed.tech/tags/channel.md>), [communication](<https://devfeed.tech/tags/communication.md>), [competition](<https://devfeed.tech/tags/competition.md>), [coupling](<https://devfeed.tech/tags/coupling.md>), [cs-ni](<https://devfeed.tech/tags/cs-ni.md>), [decentralized](<https://devfeed.tech/tags/decentralized.md>), [delivery](<https://devfeed.tech/tags/delivery.md>), [distributed](<https://devfeed.tech/tags/distributed.md>), [evolution](<https://devfeed.tech/tags/evolution.md>), [execution](<https://devfeed.tech/tags/execution.md>), [experience](<https://devfeed.tech/tags/experience.md>), [feedback](<https://devfeed.tech/tags/feedback.md>), [learning](<https://devfeed.tech/tags/learning.md>), [local](<https://devfeed.tech/tags/local.md>), [multi-agent](<https://devfeed.tech/tags/multi-agent.md>), [network](<https://devfeed.tech/tags/network.md>), [networks](<https://devfeed.tech/tags/networks.md>), [observable](<https://devfeed.tech/tags/observable.md>), [optimization](<https://devfeed.tech/tags/optimization.md>), [radio](<https://devfeed.tech/tags/radio.md>), [reinforcement-learning](<https://devfeed.tech/tags/reinforcement-learning.md>)

### AI overview

The paper proposes FAR-MAPPO for distributed, deadline-sensitive access in cognitive radio networks, where transmitters have only partial reception information and must account for interference, queue changes, and protocol overhead. Using centralized training and decentralized execution, it combines channel-specific feedback from completed access attempts with current sensing and packet urgency in a residual policy for local access decisions. The abstract reports better timely delivery than baseline methods, especially under heavy traffic, bursty arrivals, and mixed packet deadlines.

### Source excerpt

arXiv:2611.19229v1 Announce Type: new Abstract: Supporting time-sensitive services in cognitive radio networks requires distributed access strategies that translate shared spectrum opportunities into timely packet delivery. Achieving this objective is challenging because reception conditions are only partially observable to individual transmitters, while dynamic competition and access overhead constrain the service that can be completed within packet deadlines. We formulate a distributed access problem that captures the coupling among interference-dependent reception, queue evolution, and protocol durations. To address this problem, we propose feedback-aware residual multi-agent proximal policy optimization (FAR-MAPPO) under centralized training and decentralized execution. The method constructs channel-specific historical feedback representations that complement instantaneous sensing with evidence from completed access attempts. A residual policy network integrates this experience with current observations and packet urgency to guide local access decisions. Numerical results show improved timely delivery over the baseline methods, particularly under heavy traffic and bursty arrivals with mixed packet deadlines.

## Remember, Replay, Adapt: Learning from Past Experience in Online Vision-and-Language Navigation

DevFeed: [Remember, Replay, Adapt: Learning from Past Experience in Online Vision-and-Language Navigation](<https://devfeed.tech/articles/remember-replay-adapt-learning-from-past-experience-in-online-vision-and-language-navigation-137511.md>)

Original publisher: [Read original article](<https://arxiv.org/abs/2611.19901>)

Author: Chenhao Ding, Jizhou Han, Lixin Xiu, Muyao Wang, Songlin Dong, Hideki Nakayama

Published: 2026-11-17T05:00:00Z

Content type: article

Language: en

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

Topics: [Machine learning](<https://devfeed.tech/topics/machine-learning.md>), [Robot Navigation](<https://devfeed.tech/topics/robot-navigation.md>), [Computer vision](<https://devfeed.tech/topics/computer-vision.md>), [generalization in machine learning](<https://devfeed.tech/topics/generalization-in-machine-learning.md>)

Tags: [benchmarks](<https://devfeed.tech/tags/benchmarks.md>), [cs-ai](<https://devfeed.tech/tags/cs-ai.md>), [cs-cv](<https://devfeed.tech/tags/cs-cv.md>), [episodic-memory](<https://devfeed.tech/tags/episodic-memory.md>), [experience](<https://devfeed.tech/tags/experience.md>), [feedback](<https://devfeed.tech/tags/feedback.md>), [framework](<https://devfeed.tech/tags/framework.md>), [learning](<https://devfeed.tech/tags/learning.md>), [memory](<https://devfeed.tech/tags/memory.md>), [navigation](<https://devfeed.tech/tags/navigation.md>), [online](<https://devfeed.tech/tags/online.md>), [performance](<https://devfeed.tech/tags/performance.md>), [policy](<https://devfeed.tech/tags/policy.md>), [reuse](<https://devfeed.tech/tags/reuse.md>), [stable](<https://devfeed.tech/tags/stable.md>), [vision](<https://devfeed.tech/tags/vision.md>)

### AI overview

R2A adapts vision-and-language navigation policies by reusing feedback from both successful and failed past episodes. Its Success-Failure Memory retrieves relevant episodes using incoming instructions; Prior-Anchored Adaptation reinforces successful trajectories and suppresses failures relative to a frozen pretrained policy while regularizing behavioral drift. The approach is designed to work without source-domain data or expert supervision. The abstract reports consistent gains across REVERIE, R2R, and R2R-CE with five backbones; on REVERIE's validation unseen split, DUET's success rate rises from 46.98% to 69.84% and SPL from 33.73% to 50.98%.

### Source excerpt

arXiv:2611.19901v1 Announce Type: cross Abstract: Existing feedback-based methods for online vision-and-language navigation (VLN) primarily learn from current episode outcomes, leaving previously acquired feedback underutilized. Inspired by human episodic memory, we propose R2A, an online VLN adaptation framework that continually reuses successful and failed navigation experiences for policy learning. Specifically, Success-Failure Memory (SFM) retains successful and failed navigation episodes with acquired feedback and retrieves task-relevant experiences based on incoming instructions, enabling historical feedback to guide subsequent navigation. Prior-Anchored Adaptation (PAA) learns from these experiences by reinforcing successful trajectories and suppressing failed ones relative to a frozen pretrained policy, while regularizing behavioral drift. Together, these modules enable continual feedback reuse and stable policy adaptation without source-domain data or expert supervision. Extensive experiments across three benchmarks, REVERIE, R2R, and R2R-CE, and five navigation backbones demonstrate consistent performance gains. On the REVERIE validation unseen split, R2A improves DUET's SR from 46.98% to 69.84% and SPL from 33.73% to 50.98%.

## Creating a Halloween Escape Room Hotline for Twilio

DevFeed: [Creating a Halloween Escape Room Hotline for Twilio](<https://devfeed.tech/articles/creating-a-halloween-escape-room-hotline-for-twilio-142937.md>)

Original publisher: [Read original article](<https://www.twilio.com/en-us/blog/developers/halloween-escape-room-hotline-2026-twilio>)

Author: Amanda Lange

Published: 2026-11-17T00:00:00Z

Content type: tutorial

Language: en

Sources: [Twilio](<https://devfeed.tech/sources/twilio-blog.md>), [Twilio Blog](<https://devfeed.tech/sources/twilio-blog-2.md>)

Topics: [Application Development](<https://devfeed.tech/topics/application-development.md>), [twilio api](<https://devfeed.tech/topics/twilio-api.md>), [AI interaction best practices](<https://devfeed.tech/topics/ai-interaction-best-practices.md>), [Tool](<https://devfeed.tech/topics/tool.md>)

Tags: [ai](<https://devfeed.tech/tags/ai.md>), [ai-api](<https://devfeed.tech/tags/ai-api.md>), [ai-assistant](<https://devfeed.tech/tags/ai-assistant.md>), [ai-hallucinations](<https://devfeed.tech/tags/ai-hallucinations.md>), [amazon-polly](<https://devfeed.tech/tags/amazon-polly.md>), [api](<https://devfeed.tech/tags/api.md>), [azure](<https://devfeed.tech/tags/azure.md>), [blob-storage](<https://devfeed.tech/tags/blob-storage.md>), [c-sharp](<https://devfeed.tech/tags/c-sharp.md>), [claude](<https://devfeed.tech/tags/claude.md>), [claude-code](<https://devfeed.tech/tags/claude-code.md>), [coding](<https://devfeed.tech/tags/coding.md>), [developer-insights](<https://devfeed.tech/tags/developer-insights.md>), [experience](<https://devfeed.tech/tags/experience.md>), [halloween](<https://devfeed.tech/tags/halloween.md>), [ivr](<https://devfeed.tech/tags/ivr.md>), [room](<https://devfeed.tech/tags/room.md>), [tips](<https://devfeed.tech/tags/tips.md>), [tips-and-tricks](<https://devfeed.tech/tags/tips-and-tricks.md>), [twilio](<https://devfeed.tech/tags/twilio.md>), [vibe-coding](<https://devfeed.tech/tags/vibe-coding.md>)

### AI overview

The hotline combines a keypad driven IVR and SMS escape room in which players cancel a gym membership and find a door code within ten minutes. Built in C# with Twilio webhooks and TwiML Gather menus, the game uses scripted dialogue and puzzles rather than live AI, keeping outcomes consistent. The SMS version presents choices as text and lets players scroll back to clues; US automated messaging requires 10DLC certification. Testing led to shorter four digit codes and a reduced locker room puzzle. Voice lines were later pre-rendered as ElevenLabs MP3s stored in Azure blob storage, while the timer still uses Polly text to speech. Claude Code assisted with development and a script for batch voice generation, alongside manual edits. The author recommends grounding AI assisted coding in programming fundamentals, checking suggestions, and keeping prompts focused.

### Source excerpt

Creating a Halloween Escape Room Hotline for Twilio - how I did it and tips and tricks for vibe coding your own IVR escape room experience

## TokenEconomizer: Learning from Experience to Make AI Coding Agents More Efficient

DevFeed: [TokenEconomizer: Learning from Experience to Make AI Coding Agents More Efficient](<https://devfeed.tech/articles/tokeneconomizer-learning-from-experience-to-make-ai-coding-agents-more-efficient-136424.md>)

Original publisher: [Read original article](<https://www.microsoft.com/en-us/research/blog/tokeneconomizer-learning-from-experience-to-make-ai-coding-agents-more-efficient/>)

Author: Aaron J. Li, Weijia Xu, Chandan Singh, Jeevana Priya Inala, Bin Yu, Ion Stoica, Michel Galley, Jianfeng Gao

Published: 2026-11-16T17:01:35Z

Content type: article

Language: en

Sources: [Microsoft Research](<https://devfeed.tech/sources/microsoft-research.md>)

Topics: [agentic-coding](<https://devfeed.tech/topics/agentic-coding.md>), [AI-assisted coding](<https://devfeed.tech/topics/ai-assisted-coding.md>), [Machine learning](<https://devfeed.tech/topics/machine-learning.md>)

Tags: [adapting](<https://devfeed.tech/tags/adapting.md>), [agents](<https://devfeed.tech/tags/agents.md>), [ai](<https://devfeed.tech/tags/ai.md>), [ai-coding](<https://devfeed.tech/tags/ai-coding.md>), [ai-coding-agents](<https://devfeed.tech/tags/ai-coding-agents.md>), [coding](<https://devfeed.tech/tags/coding.md>), [coding-agents](<https://devfeed.tech/tags/coding-agents.md>), [costs](<https://devfeed.tech/tags/costs.md>), [experience](<https://devfeed.tech/tags/experience.md>), [learn](<https://devfeed.tech/tags/learn.md>), [learning](<https://devfeed.tech/tags/learning.md>), [microsoft](<https://devfeed.tech/tags/microsoft.md>), [microsoft-research](<https://devfeed.tech/tags/microsoft-research.md>), [open-source](<https://devfeed.tech/tags/open-source.md>), [plugin](<https://devfeed.tech/tags/plugin.md>), [research](<https://devfeed.tech/tags/research.md>), [research-blog](<https://devfeed.tech/tags/research-blog.md>), [sessions](<https://devfeed.tech/tags/sessions.md>), [tasks](<https://devfeed.tech/tags/tasks.md>), [tokens](<https://devfeed.tech/tags/tokens.md>), [waste](<https://devfeed.tech/tags/waste.md>)

### AI overview

TokenEconomizer uses repository session history to reduce coding-agent costs without changing the agent or its model. It adds confirmed recurring patterns as short startup notes, prunes tools unlikely to be needed while retaining command and file-editing tools, and uses a judge model to detect repeated failures and suggest another approach. Across 277 SWE-Bench tasks in six repositories and five agent/model combinations, costs fell 18.4-30.9% with task success broadly stable; savings included the plugin's own model calls. Early tasks could cost more while it learns, but cumulative costs were below the default by task six in every repository. The excerpt also reports savings on sequential web development and security auditing, where each task used a new repository. The plugin keeps learned data on the developer's machine and requires no API key of its own.

### Source excerpt

AI coding agents can waste time and tokens relearning the same lessons. TokenEconomizer is an open-source plugin that builds on past sessions to cut costs while maintaining task success. It's a path toward coding agents that learn not only how to solve tasks, but also how to spend computation more effectively. The post TokenEconomizer: Learning from Experience to Make AI Coding Agents More Efficient appeared first on Microsoft Research.

[Next page](<https://devfeed.tech/tags/experience.md?cursor=WyIyMDI2LTExLTE2VDE3OjAxOjM1KzAwOjAwIiwgIjNkNGQ5YjJjLTU5NzAtNDIzZi1iZmNmLWZlODZkMmIyN2Q3MyJd>)