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 | 1mo | okay this is now full stop my favorite way of describing ai labs lmao | okay this is now full stop my favorite way of describing ai labs lmao | ↗Not briefed |
arxiv arxiv-ai | 1mo | FedDermaSeg: Federated Learning for Dermatological Image Segmentation | Skin cancer is a major global health concern, and early detection and accurate lesion delineation are important for effective diagnosi… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Wikipedia has claimed a "rogue AI" operated by OpenAI has been attempting incursions on the free encyclopedia, and was even responsible for… | Wikipedia has claimed a "rogue AI" operated by OpenAI has been attempting incursions on the free encyclopedia, and was even responsibl… | ↗Not briefed |
arxiv arxiv-ai | 1mo | RAG-PIBench: A Leakage-Aware Benchmark for Prompt-Injection Detection in Trustworthy RAG Systems | Retrieval-Augmented Generation (RAG) systems are vulnerable to prompt-injection attacks embedded in retrieved content. We introduce RA… | ↗Not briefed |
arxiv arxiv-ai | 1mo | Less Is More: A Leakage-Controlled Study of Dermoscopic Preprocessing for Joint Skin Lesion Classification and Segmentation with YOLO26 | Handcrafted preprocessing is widely employed in automated dermoscopic analysis to suppress imaging artifacts and enhance lesion visibi… | ↗Not briefed |
bluesky bluesky-ai | 1mo | skill issue and, more importantly, model quality issue. these projects only fall apart because the model lets them, and when model capabili… | skill issue and, more importantly, model quality issue. these projects only fall apart because the model lets them, and when model cap… | ↗Not briefed |
bluesky bluesky-ai | 1mo | I fully understand this, as I was disheartened to see someone put an MZ emulator together using Claude. I really don't get why anyone would… | I fully understand this, as I was disheartened to see someone put an MZ emulator together using Claude. I really don't get why anyone… | ↗Not briefed |
arxiv arxiv-ai | 1mo | Singular Value Decomposition: A Geometric Rediscovery, Where Proofs Become Algorithms | This article is a geometric rediscovery of the singular value decomposition, with a further claim: the construction it builds is the m… | ↗Not briefed |
arxiv arxiv-ai | 1mo | Valid for Free: Homophily-Gated Conformal Prediction for Training-Free Node Classification with Tabular Foundation Models | Tabular foundation models (TFMs) can classify the nodes of a graph without training on it, by reading node and neighborhood features a… | ↗Not briefed |
arxiv arxiv-ai | 1mo | Adaptive Power Sampling for LLM Reasoning | Sequence-level power sampling has recently emerged as a training-free approach to reasoning by sampling from a sharpened output distri… | ↗Not briefed |
arxiv arxiv-ai | 1mo | Reinforcement Learning for Hierarchical Reasoning Rewards: Minimax-Optimal Rates with Transformers | Reinforcement learning (RL) has become a standard tool for post-training language models on reasoning tasks, where the policy is updat… | ↗Not briefed |
arxiv arxiv-ai | 1mo | Have I Seen Enough? Frozen Video-Language Models Encode Evidence Readiness | Streaming video-language models must decide not only what to answer, but whether the evidence needed for the current question has arri… | ↗Not briefed |
arxiv arxiv-ai | 1mo | Latent space bias directions in LLMs capture confidence, not fairness | Activation steering has gained popularity as a lightweight inference-time debiasing technique for large language models. However, prio… | ↗Not briefed |
hn hn-ai | 1mo | Show HN: Fill Gates – The daily logic circuit puzzle | Please try my new puzzle game. You change the inputs in a circuit full of logic gates to try and reach the target output. I got the id… | ↗Not briefed |
hn hn-ai | 1mo | Manifest-Loop | — | ↗Not briefed |
bluesky bluesky-ai | 1mo | Hmmm... are they sandbagging everyone else? GLM-5.3: Zhipu reports 66,9%, using their native agent pipeline DeepSeek V4 Pro (0813): DeepSe…🔧 Qwen 3 | Hmmm... are they sandbagging everyone else? GLM-5.3: Zhipu reports 66,9%, using their native agent pipeline DeepSeek V4 Pro (0813): D… | ↗Not briefed |
arxiv arxiv-ai | 1mo | Systemization of Knowledge (SoK): Human-Centered AI Safety for Youth | While HCI increasingly examines AI-safety for youth, the literature lacks a comprehensive view of what risks have been identified, how… | ↗Not briefed |
arxiv arxiv-ai | 1mo | AnyBottle: A Recipe to Only Keep the Concepts You Really Need | Concept bottleneck models (CBMs) make predictions inspectable and intervenable by routing them through human-interpretable concepts, b… | ↗Not briefed |
arxiv arxiv-ai | 1mo | DeltaTTT: Layerwise Optimization for Nonlinear Recurrent Memory | Sequential test-time training adapts a memory network through successive updates, each computing an inner-loop gradient based on the n… | ↗Not briefed |
hn hn-ai | 1mo | Show HN: Submilli – runtime with semantic permissions for agents that write code | — | ↗Not briefed |
hn hn-ai | 1mo | SoloMD: Open-Source Markdown Editor for macOS, Windows, Linux, Android, and iOS | — | ↗Not briefed |
hn hn-ai | 1mo | Claude Code in the Cloud | — | ↗Not briefed |
bluesky bluesky-ai | 1mo | Lads, corporate AI spend has already peaked. Look how much Anthropic (in brown) is losing already, no wonder their comms are getting despe… | Lads, corporate AI spend has already peaked. Look how much Anthropic (in brown) is losing already, no wonder their comms are getting… | ↗Not briefed |
bluesky bluesky-ai | 1mo | “Incidents like this one, and the many others that have been (and are still being) uncovered, illustrate how AI agents can drain resources … | “Incidents like this one, and the many others that have been (and are still being) uncovered, illustrate how AI agents can drain resou… | ↗Not briefed |
arxiv arxiv-ai | 1mo | How High Is 0.6? Floors, Ceilings, and Headroom in Interpretability Probing | Probes are the workhorse of interpretability. If a model's hidden states predict a variable, the model is said to represent it. But a… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Hey, @johnrogers.bsky.social How about a Leverage: Redemption episode where the crew arranges for a team of septic tank service trucks to… | Hey, @johnrogers.bsky.social How about a Leverage: Redemption episode where the crew arranges for a team of septic tank service truc… | ↗Not briefed |
bluesky bluesky-ai | 1mo | 🚀 BlueSky trending hashtags (15m) #16-30: #sexy #realnsfw #pussy #porn #model #nature #gamedev #18+ #digitalart #pubes #sex #nude #inktob… | 🚀 BlueSky trending hashtags (15m) #16-30: #sexy #realnsfw #pussy #porn #model #nature #gamedev #18+ #digitalart #pubes #sex #nude #i… | ↗Not briefed |
bluesky bluesky-ai | 1mo | 'Six scientific publishers will share more than 20 million research papers with a clinical artificial intelligence start-up, on the conditi… | 'Six scientific publishers will share more than 20 million research papers with a clinical artificial intelligence start-up, on the co… | ↗Not briefed |
arxiv arxiv-ai | 1mo | Micro Neural Policies for Safe Real-Time Robotic Control | In this paper, we investigate the synthesis of Micro Neural Policies (MNP) to enable safe and robust real-time robotic control on comp… | ↗Not briefed |
arxiv arxiv-ai | 1mo | Toward Alignment Scaling Laws: A Framework and First Preregistered Measurements | Whether alignment gets easier or harder as models grow is often argued from isolated findings, as if alignment were one property. We t… | ↗Not briefed |
bluesky bluesky-ai | 1mo | As I've said before, corporate actors like OpenAI are regularly taking down open research and publishing infrastructures – journals, books,… | As I've said before, corporate actors like OpenAI are regularly taking down open research and publishing infrastructures – journals, b… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Hugging Face事件を受けての論争があんな実質ハルシネーション的暴走を許した、いやむしろ暗示的に積極推奨したOpenAIの責任……ではなく「LLMに自我が発生したか？！」みたいな方向にばかり行ってるのもしてやられるよなぁ、という印象┐(´~`;)┌🔧 Hugging Face | Hugging Face事件を受けての論争があんな実質ハルシネーション的暴走を許した、いやむしろ暗示的に積極推奨したOpenAIの責任……ではなく「LLMに自我が発生したか？！」みたいな方向にばかり行ってるのもしてやられるよなぁ、という印象┐(´~`;)┌ | ↗Not briefed |
hn hn-ai | 1mo | Nightmare: Intro to binary exploitation / reverse engineering course | — | ↗Not briefed |
arxiv arxiv-ai | 1mo | From Shared Demand Patterns to Local Uncertainty: Probabilistic Load Forecasting by Mixing Compact Adaptations | Probabilistic load forecasting has been widely studied for power-system operation and planning, but customer- and transformer-level fo… | ↗Not briefed |
arxiv arxiv-ai | 1mo | FlowCF: Sparse Counterfactual Explanations for Mixed-Type Tabular Data using Flow Matching | In the field of Explainable AI (XAI), counterfactual (CF) explanations interpret a model's decision by suggesting the changes to the i… | ↗Not briefed |
bluesky bluesky-ai | 1mo | "I put false quotes in my book, didn't check any of them, and put my name on it. Why aren't people yelling at OpenAI and Anthropic about it… | "I put false quotes in my book, didn't check any of them, and put my name on it. Why aren't people yelling at OpenAI and Anthropic abo… | ↗Not briefed |
arxiv arxiv-ai | 1mo | How Bregman Divergences Shape Shampoo | Understanding the principles behind Shampoo has recently guided the development of more effective neural network optimizers. These met… | ↗Not briefed |
arxiv arxiv-ai | 1mo | Beyond Perturbation Magnitude: Direction-Dependent Responses in Multimodal Geometric Representations | Geometric alignment scores based on Gram determinants provide a compact way to model higher-order consistency among modalities, yet ho… | ↗Not briefed |
rss techmeme | 1mo | Vinci, which makes software used to simulate elements of chip and other hardware design, raised $250M at a $1.5B valuation (Max A. Cherney/Reuters) | Max A. Cherney / Reuters: Vinci, which makes software used to simulate elements of chip and other hardware design, raised $250M at a $… | ↗Not briefed |
rss techcrunch-ai | 1mo | LibreOffice says ‘no AI’ is now a software feature | The maker of the open source document editor says it has no plans to add AI to its software's default configuration, citing user priva… | ↗Not briefed |
bluesky bluesky-ai | 1mo | OpenAI should be asked to pay compensation to the Wikimedia Foundation. | OpenAI should be asked to pay compensation to the Wikimedia Foundation. | ↗Not briefed |
bluesky bluesky-ai | 1mo | www.commondreams.org/news/sanders... “Mr. Altman: What ‘bad things’ should we accept to advance AI?” Sanders (I-Vt.) wrote in response to t… | www.commondreams.org/news/sanders... “Mr. Altman: What ‘bad things’ should we accept to advance AI?” Sanders (I-Vt.) wrote in response… | ↗Not briefed |
arxiv arxiv-ai | 1mo | MedCORE: Criteria-Grounded Clinical Reasoning for Interpretable Medical Image Diagnosis | Clinical diagnosis is inherently a structured reasoning process, yet existing deep learning models often bypass this structure by mapp… | ↗Not briefed |
arxiv arxiv-ai | 1mo | PHBA: Prefix-State Hybrid Block Attention | Hybrid architectures combining linear sequence models with softmax attention provide an effective balance between efficient long-conte… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Every one of these leaks is a violation of the securities laws, just as SpaceX's leaks were before, and we'll call this de-regulation by no… | Every one of these leaks is a violation of the securities laws, just as SpaceX's leaks were before, and we'll call this de-regulation… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Mistral vient de sortir sa nouvelle LLM Mistral Large 4. Et donner tort à tous ceux qui disaient qu'ils avaient abandonné. Quand il sera …🔧 Mistral Large🔧 ChatGPT | Mistral vient de sortir sa nouvelle LLM Mistral Large 4. Et donner tort à tous ceux qui disaient qu'ils avaient abandonné. Quand il… | ↗Not briefed |
bluesky bluesky-ai | 1mo | ‘Le Marquer said an AI model “swarm” — in which autonomous agents work together to cause a hack, outside the control of humans — could also… | ‘Le Marquer said an AI model “swarm” — in which autonomous agents work together to cause a hack, outside the control of humans — could… | ↗Not briefed |
hn hn-ai | 1mo | Cubesat Space Protocol | — | ↗Not briefed |
arxiv arxiv-ai | 1mo | How Much Evidence Should a Coding Agent's Self-Correction Carry? Adaptive Dirichlet Evidence for Self-Distillation | Execution feedback lets coding agents revise programs and learn from their own corrections. A correction's learning weight should refl… | ↗Not briefed |
arxiv arxiv-ai | 1mo | Wiki-Talkie: Multilingual Benchmarking of Persona-Based Agents on Real-World Discussions | LLMs are increasingly deployed as autonomous agents in social environments, making it critical to study their ability to faithfully si… | ↗Not briefed |