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 |
|---|---|---|---|---|
hn hn-ai | 19d | Momentum: Luxon's API, running on Temporal | — | ↗Not briefed |
arxiv arxiv-ai | 19d | AfriGuard: Safety and Robust Refusal For African Languages | Large language models (LLMs) designed to interact with users are typically trained to respond appropriately to harmful or toxic conten… | ↗Not briefed |
bluesky bluesky-ai | 19d | LLM: Lazy Loser Machine | LLM: Lazy Loser Machine | ↗Not briefed |
arxiv arxiv-ai | 19d | A Simple and Unified Method for Gradient-Variation Regret | Gradient-variation online learning has attracted increasing attention for its connections to game theory and optimization. It has been… | ↗Not briefed |
hn hn-ai | 19d | CutefishOS Is Back with a Qt-Free, GTK-Free Vibe-Coded Desktop | — | ↗Not briefed |
arxiv arxiv-ai | 19d | On Learning Expand-and-Sparsify Representations | Expand-and-sparsify ($\texttt{EaS}$) representations lift an input to a much higher dimension through a linear map and keep only the $… | ↗Not briefed |
arxiv arxiv-ai | 19d | VeriCUA: Learning Proactive Verification for Reliable GUI Agent Rewards🔧 GPT-5 | Reinforcement learning for computer-use agents increasingly relies on proactive reward agents that interact with the post-execution GU… | ↗Not briefed |
bluesky bluesky-ai | 19d | That piece is here, also with rebuttals. If you wanted to see what the latest Deepseek model from the weekend can do for 60 cents (compare… | That piece is here, also with rebuttals. If you wanted to see what the latest Deepseek model from the weekend can do for 60 cents (co… | ↗Not briefed |
bluesky bluesky-ai | 19d | Anthropic proposes License to Kill 49 Americans whenever they want. Lobby $$$ to follow. | Anthropic proposes License to Kill 49 Americans whenever they want. Lobby $$$ to follow. | ↗Not briefed |
bluesky bluesky-ai | 19d | "Por incrível que pareça, os resultados matemáticos têm de ser aceitos pela comunidade. Não há nenhum órgão que os aprove. É uma coisa comu… | "Por incrível que pareça, os resultados matemáticos têm de ser aceitos pela comunidade. Não há nenhum órgão que os aprove. É uma coisa… | ↗Not briefed |
bluesky bluesky-ai | 19d | Source: Anthropic is preparing a model with superior intuition and reasoning that can discern unstated intentions and is better at thinking… | Source: Anthropic is preparing a model with superior intuition and reasoning that can discern unstated intentions and is better at thi… | ↗Not briefed |
hn hn-ai | 19d | Fishfolk/bones: An easy-to-use game engine for making real games | — | ↗Not briefed |
rss techmeme | 19d | Source: Anthropic is preparing a model with superior intuition and reasoning that can discern unstated intentions and is better at thinking outside of the box (Stephanie Palazzolo/The Information) | Stephanie Palazzolo / The Information: Source: Anthropic is preparing a model with superior intuition and reasoning that can discern u… | ↗Not briefed |
hn hn-ai | 19d | Evaluating an AI sales agent on 100 sites/ 15M visitors (+10% rev per visitor) | — | ↗Not briefed |
hn hn-ai | 19d | What China Gets About AI That America Doesn't [video] | — | ↗Not briefed |
arxiv arxiv-ai | 19d | Training Negatives Still Matter After Calibrating Answerability Readouts | Answerability readouts predict whether a passage supplies an answer, but whether calibration can compensate for removing negative pass… | ↗Not briefed |
hn hn-ai | 19d | Claude Audits Quantum Gravity | — | ↗Not briefed |
arxiv arxiv-ai | 19d | Residual-Based Adaptive Refinement in Hybrid Variational-Collocation PINNs for Singular Elliptic PDEs with Mixed Dirichlet-Neumann Boundary Conditions | Solutions of second-order elliptic PDE problems on nonconvex or cracked domains with mixed Dirichlet-Neumann conditions typically lie… | ↗Not briefed |
arxiv arxiv-ai | 19d | Regime-Aware Interpretable Ranking for Sector Rotation | Sector rotation is a common decision in benchmark-relative portfolio management, where portfolio managers rank sectors for overweight… | ↗Not briefed |
arxiv arxiv-ai | 19d | ALIGN: Aligning Condition Estimation and Expression in Advantage-Conditioned Behavior Cloning | Human-collected robot manipulation data often mix successful behavior with retries, pauses, recoveries, and failures. Advantage-Condit… | ↗Not briefed |
bluesky bluesky-ai | 19d | Valentina Studio & Server 18.3 are out! Cancel running queries (ValentinaDB/SQLite/DuckDB) + new AI providers: GitHub Copilot & OpenAI Code… | Valentina Studio & Server 18.3 are out! Cancel running queries (ValentinaDB/SQLite/DuckDB) + new AI providers: GitHub Copilot & OpenAI… | ↗Not briefed |
arxiv arxiv-ai | 19d | Remembering Has a Price: Learning Memory Worth under Resource Constraints | Lifelong agents continually accumulate and reuse experience to improve after deployment, yet in practice they often operate under tigh… | ↗Not briefed |
bluesky bluesky-ai | 19d | That's it, I only need the final confirmation from M$ that I have terminated my agreement with them and the reason for termination is: Thei… | That's it, I only need the final confirmation from M$ that I have terminated my agreement with them and the reason for termination is:… | ↗Not briefed |
bluesky bluesky-ai | 19d | There's 2 guys behind me in my calc class that littarely talk about AI all the fucking time in class and not just your standard bitching ab… | There's 2 guys behind me in my calc class that littarely talk about AI all the fucking time in class and not just your standard bitchi… | ↗Not briefed |
bluesky bluesky-ai | 19d | And now? Now praise is common...because it's worthless. Because people gave up entirely, and outsourced it to a machine. They went to the … | And now? Now praise is common...because it's worthless. Because people gave up entirely, and outsourced it to a machine. They went to… | ↗Not briefed |
arxiv arxiv-ai | 19d | TeleASR-Amharic: A Benchmark for Robust Amharic ASR in Realistic Telecommunication Conditions | The real-world usability of ASR models remains challenging, particularly for low-resource languages, where existing speech resources a… | ↗Not briefed |
arxiv arxiv-ai | 19d | Label Authority in Speech Emotion Recognition: Intended and Perceived Ground Truth in Emozionalmente | Ground truth in Speech Emotion Recognition (SER) is often presented as a single categorical fact, yet an acted emotion and the emotion… | ↗Not briefed |
arxiv arxiv-ai | 19d | WenSight-Decision-1 for Direct Decisions under E-commerce Rules | E-commerce decisions require applying natural-language rules to the available case evidence. We introduce WenSight-Decision-1, a share… | ↗Not briefed |
bluesky bluesky-ai | 19d | AI boys' doom shit is an exercise in 1) valuation & 2) regulatory capture. Appalling: setting thresholds for $ loss (first) & human casualt… | AI boys' doom shit is an exercise in 1) valuation & 2) regulatory capture. Appalling: setting thresholds for $ loss (first) & human ca… | ↗Not briefed |
hn hn-ai | 19d | Local OpenRouter proxy that picks a provider by cost and speed🔧 OpenRouter | — | ↗Not briefed |
arxiv arxiv-ai | 19d | Rethinking LLM Fine-Tuning via Weight Space Reparameterization: Preserving Safety during Downstream Adaptation | Fine-tuning large language models (LLMs) on downstream tasks often weakens their previously aligned safety behavior, creating an inher… | ↗Not briefed |
arxiv arxiv-ai | 19d | Fake News, Real Crashes: Information Injection Attacks on LLM Financial Trading Agents | Large language model (LLM) agents now trade real capital using news as evidence, yet they cannot reliably tell fabricated headlines fr… | ↗Not briefed |
rss theverge-ai | 19d | A hacker broke into John Deere and SharkNinja with an army of AI agents | A hacker broke into more than 100 websites using an army of AI agents that mostly operated on their own, according to findings from th… | ↗Not briefed |
arxiv arxiv-ai | 19d | R2-GP: Variance-Calibrated Priors for Additive Gaussian Processes | We develop R2-GP, an R2D2-type prior for Gaussian process (GP) generalised additive models (GAMs) that calibrates explained variance (… | ↗Not briefed |
arxiv arxiv-ai | 19d | The Geometry of Agent Memory: Forecasting Retrieval-Bound Memory Failures, and Characterizing Its Limits | We study whether geometric relationships between conversation turns, memory texts, and queries can predict answer correctness before g… | ↗Not briefed |
arxiv arxiv-ai | 19d | Physiology-Aware Graph Attention for Early Ventricular Reentry Prediction | Sudden cardiac death mostly originates from post-infarct ventricular tachycardia. This reentry is sustained by propagation around scar… | ↗Not briefed |
hn hn-ai | 19d | Show HN: KiviDB, a faster Redis-compatible memory database and vector cache | Hi all,<p>So I have been working on this project KiviDB from February 2026. It started as a project for me to understand rust, having… | ↗Not briefed |
arxiv arxiv-ai | 19d | HarnessRL: Co-Evolving Policies with a Just-in-Time Training Harness for Open-Ended RL | Rubric-based reinforcement learning (RL) trains policies for open-ended generation with natural-language criteria as rewards, but thes… | ↗Not briefed |
arxiv arxiv-ai | 19d | SNO-PI: Label-Scarce SNOMED CT Entity Linking with Pre-trained Encoders and Curated Terminology Knowledge | We present SNO-PI, the top-performing submission on SNOMED International's SNOMED CT Entity Linking Benchmark as of 18 October 2026. S… | ↗Not briefed |
bluesky bluesky-ai | 19d | China's draft export rules would cover AI model weights and source code. Already-public open weights stay exempt; future releases need lice… | China's draft export rules would cover AI model weights and source code. Already-public open weights stay exempt; future releases need… | ↗Not briefed |
arxiv arxiv-ai | 19d | Governing the Rule Space. Ruliology, Computational Irreducibility and Anticipatory Governance of Quantum Technologies | The regulation of emerging technologies is often conceived as a strategy of catching up, because legislators arrive too late. Yet gath… | ↗Not briefed |
bluesky bluesky-ai | 19d | 生成AI(llm)で情報盗んだりしてる人は改心の余地はなさそうだけど | 生成AI(llm)で情報盗んだりしてる人は改心の余地はなさそうだけど | ↗Not briefed |
bluesky bluesky-ai | 19d | ここ数か月で状況がだいぶ変わったので、最新の情報でまとめ直してみました respark.jp/t/1063 | ここ数か月で状況がだいぶ変わったので、最新の情報でまとめ直してみました respark.jp/t/1063 | ↗Not briefed |
arxiv arxiv-ai | 19d | Self-Consistency of Entity Sets and Citations in a Deep Research System under Repeated Prompting | Deep research agents (DRA) are commonly used for open-ended synthesis tasks where closed-world QA benchmarks are poorly suited to char… | ↗Not briefed |
bluesky bluesky-ai | 19d | Why did the GenAI LLM cross the road? It didn't, even though it told you it did. | Why did the GenAI LLM cross the road? It didn't, even though it told you it did. | ↗Not briefed |
arxiv arxiv-ai | 19d | Local data augmentation for imbalanced binary classification | Facing imbalanced binary classification, we aim at maximizing the recall while limiting the proportion of false detections in the mino… | ↗Not briefed |
arxiv arxiv-ai | 19d | Co-Evolving Policy and Reward for Code Generation: A Mechanistic Study of Single-Adapter Drift | Reinforcement learning with verifiable rewards (RLVR) is particularly well suited to code generation because generated programs can be… | ↗Not briefed |
bluesky bluesky-ai | 19d | LLM時代、前時代的なキーボードでの文字列うちこむより、音声で自然な文章を文字起こしできる環境を作ることが結構重要な能力になっている感じがする | LLM時代、前時代的なキーボードでの文字列うちこむより、音声で自然な文章を文字起こしできる環境を作ることが結構重要な能力になっている感じがする | ↗Not briefed |
bluesky bluesky-ai | 19d | Today's #AlgebraicGeometry lecture will discuss the Jacobian conjecture. A very different situation to when I taught this material just 10 … | Today's #AlgebraicGeometry lecture will discuss the Jacobian conjecture. A very different situation to when I taught this material jus… | ↗Not briefed |
arxiv arxiv-ai | 19d | Where Does Rhetorical Distortion Live? A Multi-Platform, Multi-Language Measurement of Influencer Content🔧 Replicate | Much of the manipulation in online influencer content is not false but disproportionate: an accurate fact is framed as historic, urgen… | ↗Not briefed |