News
The raw AI news stream — everything we ingest across RSS, Hacker News, arXiv, GitHub, Bluesky, YouTube and more, newest first.
| Source | Age | Title | Summary | Status |
|---|---|---|---|---|
bluesky bluesky-ai | 18d | Anthropic is the company Tim Houston approved for data centres in Nova Scotia🤬 Anthropic sued by major Canadian news publishers for copyr… | Anthropic is the company Tim Houston approved for data centres in Nova Scotia🤬 Anthropic sued by major Canadian news publishers for… | ↗Not briefed |
arxiv arxiv-ai | 18d | ACE-Compression: An Ability-Centric Evaluation of Compressed Large Language Models | Model compression makes large language models deployable, but today's evaluation tells an incomplete story. Standard practice measures… | ↗Not briefed |
arxiv arxiv-ai | 18d | MMEnterprise-Bench: Evaluating Enterprise Agents over Evolving Multimodal Records | Enterprise agents answer questions over workplace records that evolve through team collaboration. As information is shared, corrected,… | ↗Not briefed |
arxiv arxiv-ai | 18d | Polarity Blindness in LLM Safety: A Systematic Study of Negation, Antonym Substitution, and Adversarial Framing | Negation remains a persistent challenge for language models across NLP tasks and modalities. We show that this weakness can undermine… | ↗Not briefed |
hn hn-ai | 18d | Ask HN: What do you do as a 50yo+ programmer not ready to quit? | I'm self taught, 51yo and been going through some recent health issues. I'm in the job market in a second tier city, and ee… | ↗Not briefed |
arxiv arxiv-ai | 18d | Staged Continual Adaptation of Multimodal Foundation Models for Japanese Financial Documents | Staged post-training of multimodal foundation models chains vision-language alignment, reasoning distillation, and domain tuning, yet… | ↗Not briefed |
arxiv arxiv-ai | 18d | CollabSciReM: Aligning Language Models for Scientific Conversations via Science-Aware Reward Decomposition | Large Language Models (LLMs) are increasingly used as research collaborators for hypothesis refinement, methodological planning, and t… | ↗Not briefed |
arxiv arxiv-ai | 18d | The Shape of the DPO Hyperparameter Landscape: What to Measure, How It Moves with Scale, and When Not to Transfer It | Direct preference optimisation (DPO) is widely used for language model post-training, but its performance depends strongly on the choi… | ↗Not briefed |
bluesky bluesky-ai | 18d | AI is very much, at best, the kind of thing your boss forces you to use because he paid lot money for it or corporate demanded it and it's … | AI is very much, at best, the kind of thing your boss forces you to use because he paid lot money for it or corporate demanded it and… | ↗Not briefed |
hn hn-ai | 18d | Low Level Codeing | — | ↗Not briefed |
arxiv arxiv-ai | 18d | Testing Associative Creativity in Humans and Multimodal LLMs with Difficult-to-Name Visual Stimuli | We investigate associative creativity in descriptions of abstract visual stimuli by humans and multimodal large language models (MLLMs… | ↗Not briefed |
arxiv arxiv-ai | 18d | Physical-Variable-Informed Deep Learning for Robust Diagnosis in Industrial Fluid Transport Systems | Industrial fluid transport systems are monitored through heterogeneous sensing streams, but robust fault diagnosis remains difficult b… | ↗Not briefed |
bluesky bluesky-ai | 18d | I mean he is correct that more and more industries ae using AI Just like...very few seem to enjoy it. Like i'm hearing from techie frien… | I mean he is correct that more and more industries ae using AI Just like...very few seem to enjoy it. Like i'm hearing from techie… | ↗Not briefed |
arxiv arxiv-ai | 18d | Elicitation or Expansion? Demystifying Gains from Agent Harness Optimization | Recent studies in harness optimization have shown substantial performance gains in language model agents without updating model weight… | ↗Not briefed |
arxiv arxiv-ai | 18d | Automatic Differentiation with Constraints in Forward and Inverse Problems | Approximate solutions of partial differential equations (PDEs) on unstructured grids can be represented on a graph. Numerical optimiza… | ↗Not briefed |
hn hn-ai | 18d | You Get A Code: Keep track of App Store Offer Codes | — | ↗Not briefed |
hn hn-ai | 18d | DeepSeek just open-sourced its inference engine | — | ↗Not briefed |
hn hn-ai | 18d | Show HN: Civ 1 period accurate decompilation | I've used LLMs to decompile Civ 1 and put the reconstructed C source on GitHub, along with a technical book and story.<p>The reco… | ↗Not briefed |
hn hn-ai | 18d | How to Embed Instagram Feed Widget on Website? [video] | — | ↗Not briefed |
bluesky bluesky-ai | 18d | CERN and Mistral team up for science AI CERN and Mistral have signed a statement of intent to develop a Large Physics Model, trained on bi… | CERN and Mistral team up for science AI CERN and Mistral have signed a statement of intent to develop a Large Physics Model, trained… | ↗Not briefed |
hn hn-ai | 18d | Show HN: FlexChords – chords for any YouTube song, from a model I trained | I play guitar, and chord charts online were often wrong or oversimplified for the songs I wanted to learn, so I built FlexChords. You… | ↗Not briefed |
rss mit-tech-review-ai | 18d | A visit to Samsung’s massive new semiconductor hub | Generative AI models run on powerful chips known as GPUs, which depend on stacks of specialized memory chips packaged with the process… | ↗Not briefed |
bluesky bluesky-ai | 18d | I’ve just received the eBay email ‘Updates to our User Privacy Notice’. eBay chose NOT to draw people’s attention to using your data to tra… | I’ve just received the eBay email ‘Updates to our User Privacy Notice’. eBay chose NOT to draw people’s attention to using your data t… | ↗Not briefed |
hn hn-ai | 18d | There is no official XScreenSaver GitHub repository | — | ↗Not briefed |
hn hn-ai | 18d | Open-source benchmark for long-horizon web-app coding | — | ↗Not briefed |
arxiv arxiv-ai | 18d | Enumeration-Guided Training of Neural Hamilton-Jacobi Safety Certificates | Hamilton-Jacobi (HJ) reachability analysis offers formal safety guarantees for control systems, but training accurate neural certifica… | ↗Not briefed |
arxiv arxiv-ai | 18d | Hallucinations of Large Language Models in Cross-Language Code Translation: Phenomena and Mitigation | Code translation converts source code from one programming language to another while preserving its functionality. Although Large Lang… | ↗Not briefed |
hn hn-ai | 18d | Get More Out of Claude by Hitting the Cache | — | ↗Not briefed |
arxiv arxiv-ai | 18d | VISH-ASR: Viterbi-Guided Selection of Hotwords for Speech-LLM Contextual ASR | Prompt-based contextual ASR requires selecting a compact hotword subset from large dynamic inventories, yet existing pre-selection met… | ↗Not briefed |
bluesky bluesky-ai | 18d | This is in "9. eBay and AI", click the box "+ AI model training" and the link is as below. Also available in the bottom of a locked filing … | This is in "9. eBay and AI", click the box "+ AI model training" and the link is as below. Also available in the bottom of a locked fi… | ↗Not briefed |
bluesky bluesky-ai | 18d | OpenAI once pushed to keep AI chips from China. Now it's staying quiet. politi.co/3TRuwRQ | OpenAI once pushed to keep AI chips from China. Now it's staying quiet. politi.co/3TRuwRQ | ↗Not briefed |
arxiv arxiv-ai | 18d | From Tokens to Files: Multi-Granularity Detection and Localization of AI-Involved Code | The increasing use of large language models for code generation calls for methods that distinguish purely human-written code from AI-i… | ↗Not briefed |
bluesky bluesky-ai | 18d | ｱﾀｲは前職のときにﾍｲのkaggler集めて半年間時間とリソースあげれば、世界で5本の指に入るLLMはできると思います！！！と提言したことがあるのだが、まあ聞いてもらえる訳もなく | ｱﾀｲは前職のときにﾍｲのkaggler集めて半年間時間とリソースあげれば、世界で5本の指に入るLLMはできると思います！！！と提言したことがあるのだが、まあ聞いてもらえる訳もなく | ↗Not briefed |
arxiv arxiv-ai | 18d | Physical Operating-Condition-Guided Vision-Language Matching for Cavitation Diagnosis in Industrial Pipeline Systems | Cavitation diagnosis in industrial pipeline systems is challenging because acoustic responses vary substantially with operating condit… | ↗Not briefed |
bluesky bluesky-ai | 18d | Die Mail aller geleakten Postfächer läuft über mx.afd-hosting.de und die zeigt auf dieselbe IP wie mx.alternita-hosting.de Und das ist die… | Die Mail aller geleakten Postfächer läuft über mx.afd-hosting.de und die zeigt auf dieselbe IP wie mx.alternita-hosting.de Und das is… | ↗Not briefed |
arxiv arxiv-ai | 18d | UvA-MT's Participation in the WMT26 General Translation Shared Task | This paper presents UvA-MT's submission to the WMT 2026 General Machine Translation shared task, competing in the unconstrained track… | ↗Not briefed |
arxiv arxiv-ai | 18d | LaLa-World: Scaling Latent World Models with Natural Language Actions | This paper shows that JEPA-style latent world models (LWMs) can learn across diverse domains and data sources by adopting the language… | ↗Not briefed |
arxiv arxiv-ai | 18d | SAPD: Learning from Agent Experience without Environment Interaction | Scaling the training of Large Language Model (LLM) agents typically requires a large collection of executable environments, whose cons… | ↗Not briefed |
bluesky bluesky-ai | 18d | ・足の関節に違和感があるので今日の運動は中止する。軽くストレッチをするに留めよう。 ・生成AIの許容できる使い方と絶対許せないラインが曖昧だし、日々変動する。もしかすると技術よりも、犯罪行為に利用したり、広告やスロップでインターネットを埋め尽くす人類のほうが苦手なのかも知れない… | ・足の関節に違和感があるので今日の運動は中止する。軽くストレッチをするに留めよう。 ・生成AIの許容できる使い方と絶対許せないラインが曖昧だし、日々変動する。もしかすると技術よりも、犯罪行為に利用したり、広告やスロップでインターネットを埋め尽くす人類のほうが苦手なのか… | ↗Not briefed |
arxiv arxiv-ai | 18d | SciDiagBench: A Type-Specific Benchmark for Human-Aligned Evaluation of Scientific Diagrams | Recent agents have demonstrated strong capabilities in generating images from textual prompts, motivating their utilization for scient… | ↗Not briefed |
arxiv arxiv-ai | 18d | From Generic to Dedicated: A Novel Optimizer for Online Continual Learning | Online Continual Learning (OCL) requires models to learn from non-stationary data streams under a strict single-pass constraint, makin… | ↗Not briefed |
hn hn-ai | 18d | Running Apertus: A Swiss-Made Multimodal LLM | — | ↗Not briefed |
bluesky bluesky-ai | 18d | I don't know what image-generating AI model certain people are using that is doing this, but I've noticed a thing lately where it tries to … | I don't know what image-generating AI model certain people are using that is doing this, but I've noticed a thing lately where it trie… | ↗Not briefed |
bluesky bluesky-ai | 18d | The biggest newspapers in Canada just sued Anthropic for alleged copyright infringement. The number of lawsuits piling up over this techno… | The biggest newspapers in Canada just sued Anthropic for alleged copyright infringement. The number of lawsuits piling up over this t… | ↗Not briefed |
hn hn-ai | 18d | CPU-native. Deterministic. O(1) processes. – the Old New Coolness | Billions over billions are burned in expensive GPU-processes designed to structure unstructured data and information. I’m not a techn… | ↗Not briefed |
arxiv arxiv-ai | 18d | COMPASS: Routing via Capability-Oriented Model Profiling with Adaptive Search, Selection, and Tool Use | LLM routing aims to select a capable model at low cost, but specialized routers can require updates as models and tasks change. We pro… | ↗Not briefed |
rss marktechpost | 18d | Exa 3.0 Search API Tops Perplexity, Brave and Parallel on 8 Benchmarks | Exa 3.0 is a search API upgrade from Exa. It adds a larger embedding model, Exa Lite at $4 per 1,000 requests, and Dynamic Highlights,… | ↗Not briefed |
arxiv arxiv-ai | 18d | SMART: Semantic Segment Mapping and Assessment of Recognition Transcripts with Contextual Rubrics | Word error rate (WER) assigns equal cost to lexical edits despite their different semantic consequences, while overall semantic scores… | ↗Not briefed |
arxiv arxiv-ai | 18d | Privacy-preserving inference on hosted models: a learned input-embedding transform for frozen Qwen3.8-27B | Every prompt sent to a hosted language model is readable by whoever runs it. We study whether the prompt can remain on the client whil… | ↗Not briefed |
bluesky bluesky-ai | 18d | No, it won’t. | No, it won’t. | ↗Not briefed |