# recycling

Published articles for recycling.

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

## RECAP: Context-Aware Prediction Recycling with Request-Local and Persistent Memories for Multimodal Speculative Decoding

DevFeed: [RECAP: Context-Aware Prediction Recycling with Request-Local and Persistent Memories for Multimodal Speculative Decoding](<https://devfeed.tech/articles/recap-context-aware-prediction-recycling-with-request-local-and-persistent-memories-for-multimodal-speculative-decoding-118396.md>)

Original publisher: [Read original article](<https://arxiv.org/abs/2611.05894>)

Author: Hengjie Zhu, Dayan Wu, Xinze Liu, Peng Fu, Zheng Lin, Weiping Wang

Published: 2026-11-05T05:00:00Z

Content type: article

Language: en

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

Topics: [Cache](<https://devfeed.tech/topics/cache.md>)

Tags: [autoregressive](<https://devfeed.tech/tags/autoregressive.md>), [benchmarks](<https://devfeed.tech/tags/benchmarks.md>), [context](<https://devfeed.tech/tags/context.md>), [cs-ai](<https://devfeed.tech/tags/cs-ai.md>), [local](<https://devfeed.tech/tags/local.md>), [memory](<https://devfeed.tech/tags/memory.md>), [multimodal](<https://devfeed.tech/tags/multimodal.md>), [predictions](<https://devfeed.tech/tags/predictions.md>), [recap](<https://devfeed.tech/tags/recap.md>), [recycling](<https://devfeed.tech/tags/recycling.md>), [request](<https://devfeed.tech/tags/request.md>), [reuse](<https://devfeed.tech/tags/reuse.md>), [speculative-decoding](<https://devfeed.tech/tags/speculative-decoding.md>), [speed](<https://devfeed.tech/tags/speed.md>), [target](<https://devfeed.tech/tags/target.md>), [vision-language-models](<https://devfeed.tech/tags/vision-language-models.md>)

### AI overview

RECAP reuses a vision-language model's own predictions for speculative decoding while reducing errors caused by matching only the last token across different images and prefixes. It combines one-, two-, and three-token keys: contextual memories are limited to the current request, while token-level memory persists across requests. Entropy-confidence weighting fuses retrieved predictions, and best-first search builds a candidate tree that one target-model pass verifies under greedy decoding. The method requires neither training nor an auxiliary draft model. Across seven single-turn benchmarks, it reports mean speedups of 1.66x on Qwen2.5-VL-7B and 1.56x on LLaVA-1.6-7B, compared with 1.51x and 1.37x for Token Recycling. These are abstract-level results; the supplied text does not provide benchmark details.

### Source excerpt

arXiv:2611.05894v1 Announce Type: new Abstract: Recycling a target model's own predictions offers a training-free source of proposals for speculative decoding. In vision-language models, however, existing recycling indexes predictions by the last token alone, conflating predictions computed under different images and prefixes; we find that such reuse fails most often at image-sensitive decoding states. Longer keys reduce this aliasing but match fewer states. We introduce \method, which stores target predictions under one-, two-, and three-token keys: contextual memories live only for the current request and are thus implicitly conditioned on its image, while token-level memory persists across requests to retain coverage. At each lookup, entropy-confidence weighting fuses the retrieved predictions according to their key specificity and concentration. Best-first search composes the fused transitions into a candidate tree, supplemented by suffix-matched continuations, and one target pass verifies the tree under greedy decoding. Without training or an auxiliary draft model, \method\ achieves mean speedups over autoregressive decoding of $1.66\speed$ and $1.56\speed$ across seven single-turn benchmarks on Qwen2.5-VL-7B and LLaVA-1.6-7B, exceeding the strongest training-free baseline, Token Recycling ($1.51\speed$ and $1.37\speed$). Our implementation is available in https://github.com/Senmo996/RECAP-official-code.

## The Repair Café start packet for Linux installations

DevFeed: [The Repair Café start packet for Linux installations](<https://devfeed.tech/articles/the-repair-cafe-start-packet-for-linux-installations-106241.md>)

Original publisher: [Read original article](<https://blog.adafruit.com/2026/10/29/the-repair-cafe-start-packet-for-linux-installations/>)

Author: Anne Barela

Published: 2026-10-29T12:51:33Z

Content type: news

Language: en

Sources: [Adafruit Industries](<https://devfeed.tech/sources/adafruit-industries-makers-hackers-artists-designers-and-engineers.md>)

Topics: [Linux Mint](<https://devfeed.tech/topics/linuxmint.md>), [Linux](<https://devfeed.tech/topics/linux.md>)

Tags: [anniversary](<https://devfeed.tech/tags/anniversary.md>), [computers](<https://devfeed.tech/tags/computers.md>), [free](<https://devfeed.tech/tags/free.md>), [groups](<https://devfeed.tech/tags/groups.md>), [guide](<https://devfeed.tech/tags/guide.md>), [help](<https://devfeed.tech/tags/help.md>), [install](<https://devfeed.tech/tags/install.md>), [installations](<https://devfeed.tech/tags/installations.md>), [instead](<https://devfeed.tech/tags/instead.md>), [laptop](<https://devfeed.tech/tags/laptop.md>), [laptops](<https://devfeed.tech/tags/laptops.md>), [linux](<https://devfeed.tech/tags/linux.md>), [linux-mint](<https://devfeed.tech/tags/linux-mint.md>), [microsoft](<https://devfeed.tech/tags/microsoft.md>), [netherlands](<https://devfeed.tech/tags/netherlands.md>), [network](<https://devfeed.tech/tags/network.md>), [open-source](<https://devfeed.tech/tags/open-source.md>), [operating-system](<https://devfeed.tech/tags/operating-system.md>), [recycling](<https://devfeed.tech/tags/recycling.md>), [refugees](<https://devfeed.tech/tags/refugees.md>), [repair](<https://devfeed.tech/tags/repair.md>), [repair-cafe](<https://devfeed.tech/tags/repair-cafe.md>), [windows-10](<https://devfeed.tech/tags/windows-10.md>)

### AI overview

Repair Café International's starter kit helps cafés run sessions where visitors can have Linux Mint installed on old laptops. One year after its launch, 165 Repair Cafés worldwide were offering this, addressing computers that cannot upgrade to Windows 11 as Windows 10 support approaches its end in October 2027. Cafés without installation expertise can seek help: in the Netherlands, the Nederlandse Linux Gebruikers Groep supports the effort through BuurtLinux, a volunteer network offering free post-install support. Its LaptopRevive project also installs Linux on discarded laptops and gives them to students, refugees, and others on limited budgets. The excerpt notes that similar groups operate in other countries.

### Source excerpt

Repair Café International marked the first anniversary of its Linux Repair Café starter kit on repaircafe.org. The kit is a guide for running sessions where visitors bring an old laptop and leave with Linux Mint installed. One year in, 165 Repair Cafés around the world are doing this. The reason is Windows 10. Microsoft ends [...]

## ReLEAP: Upcycling Vision Datasets through Cross-Task Unification and Semi-Supervised Annotation

DevFeed: [ReLEAP: Upcycling Vision Datasets through Cross-Task Unification and Semi-Supervised Annotation](<https://devfeed.tech/articles/releap-upcycling-vision-datasets-through-cross-task-unification-and-semi-supervised-annotation-99594.md>)

Original publisher: [Read original article](<https://arxiv.org/abs/2610.32289>)

Author: Julia Simon, Daniel Steininger, Andreas Trondl

Published: 2026-10-26T04:00:00Z

Content type: article

Language: en

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

Topics: [image recognition](<https://devfeed.tech/topics/image-recognition.md>), [Image](<https://devfeed.tech/topics/image.md>), [AI Foundation Models](<https://devfeed.tech/topics/ai-foundation-models.md>), [Fine-tuning](<https://devfeed.tech/topics/fine-tuning.md>), [image tagging](<https://devfeed.tech/topics/image-tagging.md>)

Tags: [agriculture](<https://devfeed.tech/tags/agriculture.md>), [annotation](<https://devfeed.tech/tags/annotation.md>), [cs-cv](<https://devfeed.tech/tags/cs-cv.md>), [cs-lg](<https://devfeed.tech/tags/cs-lg.md>), [datasets](<https://devfeed.tech/tags/datasets.md>), [fine-tuning](<https://devfeed.tech/tags/fine-tuning.md>), [framework](<https://devfeed.tech/tags/framework.md>), [image](<https://devfeed.tech/tags/image.md>), [learning](<https://devfeed.tech/tags/learning.md>), [models](<https://devfeed.tech/tags/models.md>), [precision](<https://devfeed.tech/tags/precision.md>), [python](<https://devfeed.tech/tags/python.md>), [recycling](<https://devfeed.tech/tags/recycling.md>), [research](<https://devfeed.tech/tags/research.md>), [reuse](<https://devfeed.tech/tags/reuse.md>), [semi](<https://devfeed.tech/tags/semi.md>), [taxonomy](<https://devfeed.tech/tags/taxonomy.md>), [training-data](<https://devfeed.tech/tags/training-data.md>), [upcycling](<https://devfeed.tech/tags/upcycling.md>), [vision](<https://devfeed.tech/tags/vision.md>)

### AI overview

ReLEAP is a Python framework for making heterogeneous vision datasets reusable across tasks. It abstracts data sources, reuses annotations across learning tasks, and maps differing labels to a shared hierarchy. In a precision agriculture demonstration, it unified 15 datasets spanning plant species and annotation types. The framework also supports annotation upcycling: experiments examine how domain-specific models trained on unified data and foundation models generate higher-level annotations under unsupervised, weakly supervised, and semi-supervised settings, including how dataset characteristics affect results. The authors report expanded pools of reusable data and annotations and improved transparency. The framework and trained models are released for academic use.

### Source excerpt

arXiv:2610.32289v1 Announce Type: cross Abstract: Developing robust vision models requires large amounts of annotated data, especially for tasks involving fine-grained class distinctions under variable environmental conditions. Even in domains with numerous existing datasets, differences in annotation formats, label definitions and targeted learning tasks hinder data reuse and integration. To address this challenge, we introduce ReLEAP, a scalable Python framework for unifying heterogeneous image datasets through data-source abstraction, systematic reuse of annotations across learning tasks, and mapping of labels to a shared hierarchical taxonomy. We demonstrate its applicability in precision agriculture by integrating fifteen diverse datasets covering a broad range of plant species and annotation types. Beyond unification, ReLEAP enables systematic recycling of existing data and provides a robust foundation for learning-based annotation upcycling. Leveraging this foundation, we conduct extensive cross-dataset and fine-tuning experiments to investigate how domain-specific models trained on unified data and foundation models can be used to automatically generate higher-level annotations under unsupervised, weakly supervised and semi-supervised settings, and how dataset characteristics influence their effectiveness. Our results show that unified data access and annotation upcycling substantially expand the pool of reusable training data and annotations while improving data transparency, enabling more sustainable and resource-efficient use of existing datasets. We release the framework and trained models for academic use to support accessible and collaborative research, particularly in data-scarce and heterogeneous domains.

## Co-Evolving Policy and Reward for Code Generation: A Mechanistic Study of Single-Adapter Drift

DevFeed: [Co-Evolving Policy and Reward for Code Generation: A Mechanistic Study of Single-Adapter Drift](<https://devfeed.tech/articles/co-evolving-policy-and-reward-for-code-generation-a-mechanistic-study-of-single-adapter-drift-93626.md>)

Original publisher: [Read original article](<https://arxiv.org/abs/2610.26416>)

Author: Amarsaikhan Batjargal, Noriko Tomuro

Published: 2026-10-21T04:00:00Z

Content type: article

Language: en

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

Topics: [rlvr](<https://devfeed.tech/topics/rlvr.md>), [grpo](<https://devfeed.tech/topics/grpo.md>), [Reinforcement learning](<https://devfeed.tech/topics/reinforcement-learning.md>)

Tags: [adapter](<https://devfeed.tech/tags/adapter.md>), [code](<https://devfeed.tech/tags/code.md>), [code-generation](<https://devfeed.tech/tags/code-generation.md>), [cs-cl](<https://devfeed.tech/tags/cs-cl.md>), [cs-lg](<https://devfeed.tech/tags/cs-lg.md>), [cs-se](<https://devfeed.tech/tags/cs-se.md>), [diagnostics](<https://devfeed.tech/tags/diagnostics.md>), [evaluation](<https://devfeed.tech/tags/evaluation.md>), [evolution](<https://devfeed.tech/tags/evolution.md>), [execution](<https://devfeed.tech/tags/execution.md>), [grpo](<https://devfeed.tech/tags/grpo.md>), [learning](<https://devfeed.tech/tags/learning.md>), [model](<https://devfeed.tech/tags/model.md>), [optimization](<https://devfeed.tech/tags/optimization.md>), [performance](<https://devfeed.tech/tags/performance.md>), [policy](<https://devfeed.tech/tags/policy.md>), [problems](<https://devfeed.tech/tags/problems.md>), [recycling](<https://devfeed.tech/tags/recycling.md>), [reflection](<https://devfeed.tech/tags/reflection.md>), [reinforcement-learning](<https://devfeed.tech/tags/reinforcement-learning.md>), [reward](<https://devfeed.tech/tags/reward.md>), [rlvr](<https://devfeed.tech/tags/rlvr.md>), [self-improvement](<https://devfeed.tech/tags/self-improvement.md>)

### AI overview

In experiments with Qwen2.5-Coder-3B-Instruct, SPARK-style rollout recycling improved training performance but did not yield sustained gains on held-out code problems, and the results did not support added benefits from execution-grounded reflection. At the same training stage, rollout recycling produced roughly 34 times the KL divergence from the reference policy seen with execution-only GRPO. As training pass rate rose, trained tokens per GRPO sequence fell from 185 to 86 while held-out performance declined or failed to sustain early gains. The authors describe this as single-adapter policy drift: shared parameters move away from pretrained code-generation behavior without corresponding generalization. They caution that higher training reward alone does not demonstrate self-improvement and emphasize held-out evaluation and drift diagnostics.

### Source excerpt

arXiv:2610.26416v1 Announce Type: cross Abstract: Reinforcement learning with verifiable rewards (RLVR) is particularly well suited to code generation because generated programs can be objectively evaluated through execution. Execution also provides behavioral information that can be reused as additional supervision. The Synergistic Policy and Reward Co-Evolving Framework (SPARK), originally proposed for reasoning tasks, exploits this idea by recycling training rollouts and their evaluation results to train a model not only to generate solutions but also to evaluate and improve them. We investigate whether this policy-reward co-evolution approach transfers effectively to code generation and whether execution-grounded reflection provides an additional source of self-improvement. We compare SPARK-style rollout recycling with execution-only Group Relative Policy Optimization (GRPO) using Qwen2.5-Coder-3B-Instruct. Our results do not support either hypothesis under the configuration tested. Although co-evolution improves training performance, these gains do not generalize to held-out problems. Compared at the same stage of training, rollout recycling produces roughly 34x greater KL divergence from the reference policy than execution-only GRPO. Over training, trained tokens per GRPO sequence fall from 185 to 86 as training pass rate increases, while held-out performance declines or fails to sustain early gains. We characterize these patterns as single-adapter policy drift: the shared trainable parameters increasingly depart from pretrained code-generation behavior without corresponding gains in generalization. The results show that higher training reward alone is not evidence of generalizable code self-improvement and highlight the importance of held-out evaluation and drift diagnostics.

## Danu Robotics' fight to build a better recycling robot

DevFeed: [Danu Robotics' fight to build a better recycling robot](<https://devfeed.tech/articles/danu-robotics-fight-to-build-a-better-recycling-robot-84885.md>)

Original publisher: [Read original article](<https://techcrunch.com/2026/10/09/danu-robotics-fight-to-build-a-better-recycling-robot/>)

Author: Russell Brandom

Published: 2026-10-09T16:45:00Z

Content type: news

Language: en

Sources: [TechCrunch | AI News & Artificial Intelligence](<https://devfeed.tech/sources/techcrunch-ai-news-artificial-intelligence.md>)

Topics: [robot sense of touch](<https://devfeed.tech/topics/robot-sense-of-touch.md>), [contact-rich manipulation](<https://devfeed.tech/topics/contact-rich-manipulation.md>), [A/B Testing](<https://devfeed.tech/topics/a-b-testing.md>)

Tags: [ai](<https://devfeed.tech/tags/ai.md>), [battlefield-200](<https://devfeed.tech/tags/battlefield-200.md>), [business](<https://devfeed.tech/tags/business.md>), [funding](<https://devfeed.tech/tags/funding.md>), [hardware](<https://devfeed.tech/tags/hardware.md>), [recycling](<https://devfeed.tech/tags/recycling.md>), [robot](<https://devfeed.tech/tags/robot.md>), [robotics](<https://devfeed.tech/tags/robotics.md>), [techcrunch-disrupt](<https://devfeed.tech/tags/techcrunch-disrupt.md>)

### AI overview

Danu Robotics, founded by former software developer Amy Ma, is developing H.E.R.O., an AI-enabled robot that sorts recyclable waste. The company says it has signed $500,000 in contracts and raised $5 million in late-seed funding, while Ma sees opportunities to make the robot smaller and usable outside dedicated recycling facilities.

### Source excerpt

For six years, Danu founder Amy Ma has been working on a better way to sort recyclable waste.

## Enhance Your Application Using Picasso

DevFeed: [Enhance Your Application Using Picasso](<https://devfeed.tech/articles/enhance-your-application-using-picasso-15627.md>)

Original publisher: [Read original article](<https://developer.squareup.com/blog/enhance-your-application-using-picasso>)

Author: Square Engineering

Published: 2013-05-14T16:03:00Z

Content type: article

Language: en

Sources: [Jake Wharton](<https://devfeed.tech/sources/jake-wharton.md>), [Square Corner](<https://devfeed.tech/sources/square-corner-blog-medium.md>)

Topics: [Picasso](<https://devfeed.tech/topics/picasso.md>), [Cache](<https://devfeed.tech/topics/cache.md>), [Android](<https://devfeed.tech/topics/android.md>), [Image](<https://devfeed.tech/topics/image.md>)

Tags: [adapter](<https://devfeed.tech/tags/adapter.md>), [android](<https://devfeed.tech/tags/android.md>), [automatically](<https://devfeed.tech/tags/automatically.md>), [caching](<https://devfeed.tech/tags/caching.md>), [engineering](<https://devfeed.tech/tags/engineering.md>), [fluent](<https://devfeed.tech/tags/fluent.md>), [image](<https://devfeed.tech/tags/image.md>), [library](<https://devfeed.tech/tags/library.md>), [picasso](<https://devfeed.tech/tags/picasso.md>), [recycling](<https://devfeed.tech/tags/recycling.md>), [square](<https://devfeed.tech/tags/square.md>), [transformation](<https://devfeed.tech/tags/transformation.md>)

### AI overview

Picasso is an Android library for downloading and displaying images with a short fluent call. It handles caching, view recycling, and placing the finished bitmap in the target view, including when loading images in an adapter. Its resize and center-crop transformations run on the background thread used to decode the source, and the transformed result is stored in memory so different versions of the same image can be reused. Picasso uses memory caching and disk caching provided by the HTTP client; a development marker can show the image source.

### Source excerpt

A fluent image downloading and caching library for Android.