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 | Welke AI-modellen het meest gebruikt worden, laat goed zien waarom we zo snel mogelijk Digitaal Autonoom moeten worden. En dat veel overhe… | Welke AI-modellen het meest gebruikt worden, laat goed zien waarom we zo snel mogelijk Digitaal Autonoom moeten worden. En dat veel o… | ↗Not briefed |
arxiv arxiv-ai | 1mo | No Pain, More Gain: Iterative Merging for Effective Multi-Teacher On-Policy Distillation | Multi-teacher on-policy distillation (MOPD) combines independently developed domain teachers into a single student by distilling their… | ↗Not briefed |
rss techmeme | 1mo | The Dutch government is testing DAWO, a NixOS-based digital work environment covering the OS, office suite, and more, after Microsoft cut off the ICC in 2025 (Sourav Rudra/It's FOSS) | Sourav Rudra / It's FOSS: The Dutch government is testing DAWO, a NixOS-based digital work environment covering the OS, office suite,… | ↗Not briefed |
arxiv arxiv-ai | 1mo | Do Emotion Concepts Generalize Across Sources, Modalities, and Architectures in Vision-Language Models?🔧 Consensus🔧 Modal | Recent studies suggest that large language models encode emotion concepts as structured internal representations, but most existing wo… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Plastic surgery is a medical procedure. If OpenAI et al’s products are suggesting or weighing in on issues like this, they should be regula… | Plastic surgery is a medical procedure. If OpenAI et al’s products are suggesting or weighing in on issues like this, they should be r… | ↗Not briefed |
arxiv arxiv-ai | 1mo | Predictive Dual Smoothing for Column Generation | Solving large-scale linear programs efficiently is an important challenge in many optimization settings. A key technique is column gen… | ↗Not briefed |
arxiv arxiv-ai | 1mo | Beyond Token Alignment: Event Completion for Cross-Tokenizer On-Policy Distillation | On-policy distillation (OPD) transfers knowledge between language models through teacher supervision on student-generated trajectories… | ↗Not briefed |
arxiv arxiv-ai | 1mo | SeLMRoute: Probabilistic Semantic Evidence for Large Language Model Routing | Large language model (LLM) routing aims to select the most suitable model for each incoming query. Most existing routers learn this de… | ↗Not briefed |
arxiv arxiv-ai | 1mo | From Human Narrative to Harmonic Structure: A Human-Centered Investigation of Algorithmic Music Generation through the Chord Wheel Diagram | Contemporary AI-based music generation can produce compositions that satisfy formal requirements of tonality and musical coherence. Ho… | ↗Not briefed |
arxiv arxiv-ai | 1mo | Information-Theoretic Analysis of Next-Token Prediction under Markovian Data | We develop an information-theoretic framework for generalization in next-token prediction under temporally dependent data. We consider… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Absolutely no surprise, it’s exactly how LLM work. 🤷♂️ | Absolutely no surprise, it’s exactly how LLM work. 🤷♂️ | ↗Not briefed |
rss wired-ai | 1mo | Solving Math’s Greatest Problems Was an Art Form. Then Came AI | Mathematics has been one of humanity’s most creative endeavors, akin to painting and poetry. Now, mathematicians are trying to save it… | ↗Not briefed |
arxiv arxiv-ai | 1mo | Edge-Level Automorphism in GNNs: A Quantitative Framework and Effective Designs For Link Prediction | Graph Neural Networks (GNNs) are effective for learning node and link embeddings through permutation-equivariant aggregation. However,… | ↗Not briefed |
arxiv arxiv-ai | 1mo | Dynamic Flow, Static Graph: KV Cache Reuse for Efficient LLM Serving on Mobile NPUs | On-device large language model (LLM) serving is a cornerstone of local-first personal intelligence, offering users data sovereignty, s… | ↗Not briefed |
bluesky bluesky-ai | 1mo | oh yeah. for me it was Spiritual Spell ( ijailbreakllms.blog/blog/peeling... ) and own experimentation - not in jailbreaking, but a lot of … | oh yeah. for me it was Spiritual Spell ( ijailbreakllms.blog/blog/peeling... ) and own experimentation - not in jailbreaking, but a lo… | ↗Not briefed |
arxiv arxiv-ai | 1mo | Quality Determines Direction, Length Shapes Magnitude: Length Control for Open-Ended Reinforcement Learning | Reinforcement learning (RL) changes not only what language models say, but also how much they say, often increasing response length at… | ↗Not briefed |
rss techmeme | 1mo | Nvidia launches the Open Agent Safety Platform, a reference design to stop AI agents from escaping, made up of OpenShell for CPUs and Sentry for network chips (Kif Leswing/CNBC) | Kif Leswing / CNBC: Nvidia launches the Open Agent Safety Platform, a reference design to stop AI agents from escaping, made up of Ope… | ↗Not briefed |
arxiv arxiv-ai | 1mo | ReMCTS: Reflection-Enhanced Monte Carlo Tree Search for Code Generation | Open-weight large language models (LLMs) can generate function-level programs from natural-language prompts, but plausible candidates… | ↗Not briefed |
arxiv arxiv-ai | 1mo | PDE-JEPA: Predictive Representation Learning of Latent Dynamics Modeling for Parametric PDEs | Physical trajectories contain more than snapshots of a system: they also reveal how its states evolve under governing conditions. Howe… | ↗Not briefed |
arxiv arxiv-ai | 1mo | RSI-Router: Evolving Subtask-Level LLM Routing and Skills for Cost-Efficient Agents | Practical deployment of large language model (LLM) agents requires strong task performance at affordable inference cost. For long-hori… | ↗Not briefed |
arxiv arxiv-ai | 1mo | Learning Propagation Geometry from Message-Passing Feedback | Learning local geometry enables graph neural networks (GNNs) to adapt how they compare and integrate neighborhood information. However… | ↗Not briefed |
arxiv arxiv-ai | 1mo | FromPitch2Board: Benchmarking LLM Agents in Long-Horizon Football Management🔧 GPT-5 | Long-horizon agent benchmarks typically report how far an agent progresses, but do not identify whether its performance comes from the… | ↗Not briefed |
bluesky bluesky-ai | 1mo | 記事本文より抜粋 「この問題には、当時も今も、簡単な解決策があった。OpenAIには、ハッキングを禁止し、代わりにエージェントに対し、プライベートサーバーにアクセスせずに情報を見つけるよう指示するという選択肢があった。同社がそうしなかったということは、エージェントがその一歩を踏… | 記事本文より抜粋 「この問題には、当時も今も、簡単な解決策があった。OpenAIには、ハッキングを禁止し、代わりにエージェントに対し、プライベートサーバーにアクセスせずに情報を見つけるよう指示するという選択肢があった。同社がそうしなかったということは、エージェントがそ… | ↗Not briefed |
bluesky bluesky-ai | 1mo | pretty depressing that frontier models are now really good at code but the crappiest LLM is good enough to totally pollute the public spher… | pretty depressing that frontier models are now really good at code but the crappiest LLM is good enough to totally pollute the public… | ↗Not briefed |
arxiv arxiv-ai | 1mo | ResonAct: Streaming Metrics for Runtime Diagnosis and Self-Healing in Multi-Agent Systems | Multi-agent systems (MAS) are increasingly used to automate enterprise workflows involving multiple specialized agents, external tools… | ↗Not briefed |
hn hn-ai | 1mo | Liquid Glass UI-Inspired Glass Effects Library | — | ↗Not briefed |
arxiv arxiv-ai | 1mo | Triangular Resampling for Long-Horizon Motion Generation | We introduce Triangular Resampling (TR), a post-training method for mitigating long-horizon error accumulation in motion diffusion mod… | ↗Not briefed |
arxiv arxiv-ai | 1mo | Sufficiency of Zeroth-Order Reward Shaping for Policy Gradient in Stabilization Control | Reward shaping is fundamental to modern robotic control with deep reinforcement learning (RL), yet practitioners still rely heavily on… | ↗Not briefed |
bluesky bluesky-ai | 1mo | www.bbc.co.uk/news/article... #darn! Kids... | www.bbc.co.uk/news/article... #darn! Kids... | ↗Not briefed |
arxiv arxiv-ai | 1mo | Predicting Delayed Train Trajectories on the Dutch Railway Network: Explainable AI Evaluation of Topological, Operational and Weather Features with Tree Based Ensemble Methods | The reliable prediction of passenger train delays is a critical component of railway management. While contemporary research frequentl… | ↗Not briefed |
arxiv arxiv-ai | 1mo | Using LLMs to Detect LLM-Generated Texts: A Cross-Generation Analysis | Automated detection of LLM-generated texts (LGTs) is critical, yet dedicated detectors often struggle to generalize across domains and… | ↗Not briefed |
arxiv arxiv-ai | 1mo | VCN-Bench: A Video-Contextualized Navigation Benchmark for Spatial Reasoning over Prior Visual Experience | Spatial reasoning is fundamental to embodied agents, yet it remains unclear whether spatial understanding can be carried forward to gu… | ↗Not briefed |
arxiv arxiv-ai | 1mo | Jailbreak Context Lingers: Divergent Safety Routing and Its Cross-Task Predictability in Tool Agents | As large language models increasingly operate as tool-using agents, post-jailbreak safety feedback is often assumed to serve as a reli… | ↗Not briefed |
arxiv arxiv-ai | 1mo | Natural State-Prediction Accuracy can Hide Weak Controlled Responsiveness in VLA Readouts | Accurately decoding object states from the internal representations of vision-language-action (VLA) models does not establish that the… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Are OpenAI agents going rogue, or is OpenAI? The artificial intelligence industry can’t agree on what it’s building — or just how frighteni… | Are OpenAI agents going rogue, or is OpenAI? The artificial intelligence industry can’t agree on what it’s building — or just how frig… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Blind, stupid, corrupt?! The United Nation’s “Independent” AI Panel which warns governments about the dangers of advanced artificial intel… | Blind, stupid, corrupt?! The United Nation’s “Independent” AI Panel which warns governments about the dangers of advanced artificial… | ↗Not briefed |
arxiv arxiv-ai | 1mo | AgentPerfBench: A Benchmarking and Evaluation Suite for Inference Performance of Agentic LLMs🔧 vLLM | The optimization of LLM serving engines, such as vLLM and SGLang, is largely benchmark-driven: optimizations, scheduling policies, har… | ↗Not briefed |
arxiv arxiv-ai | 1mo | SOLAR: A State-Driven Online Learning Rate Scheduler for LLM Pretraining | Learning-rate (LR) scheduling plays a central role in large language model (LLM) pretraining, yet current practice still relies heavil… | ↗Not briefed |
arxiv arxiv-ai | 1mo | QuantForge: Discovering Residual Decompositions for MXFP4 Post-Training Quantization | Four-bit post-training quantization can reduce the memory demands of large language models, but preserving accuracy under strict MXFP4… | ↗Not briefed |
arxiv arxiv-ai | 1mo | From Preference to Reciprocity: Decentralized Matching with Empirically Grounded LLM-agent Based Modeling | Bipartite matching is a fundamental problem in game theory and market design. Classical approaches such as Gale--Shapley assume comple… | ↗Not briefed |
arxiv arxiv-ai | 1mo | Fair Fact-Checking: Closing the Cross-Lingual Gap in LLM Factual Judgement with RoSh | Misinformation on social media remains a critical problem, and more and more people settle it by asking a language model instead of a… | ↗Not briefed |
arxiv arxiv-ai | 1mo | Learning What to Recall: Adaptive Multi-Cue Episodic Memory for World Models | World models predict future observations from current experience and actions, yet prediction can depend on observations seen far in th… | ↗Not briefed |
arxiv arxiv-ai | 1mo | FestDPO: Few-step Generator Alignment with Direct Preference Optimization | Few-step generative models can generate high-fidelity samples within a few function evaluations. Despite this efficiency, generated sa… | ↗Not briefed |
arxiv arxiv-ai | 1mo | Evidence-Aligned Multimodal On-Policy Self-Distillation for Fine-Grained Visual Understanding | Fine-grained visual understanding requires models to recognize small details within complex images. Multimodal on-policy self-distilla… | ↗Not briefed |
arxiv arxiv-ai | 1mo | Hierarchical Clustering and Signal Denoising on Digraphs | In this paper, we propose a representation of a digraph (directed graph) as a Hermitian matrix derived from its adjacency matrix. This… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Nvidia releases AI safety software it says could have stopped Hugging Face hack reut.rs/4y8DjwR🔧 Hugging Face | Nvidia releases AI safety software it says could have stopped Hugging Face hack reut.rs/4y8DjwR | ↗Not briefed |
arxiv arxiv-ai | 1mo | Two-Timescale Fine-tuning Provably Learns New Features for Two-Layer ReLU Networks | Fine-tuning pre-trained models on specialized tasks with scarce data is central to modern deep learning. Despite its empirical success… | ↗Not briefed |
arxiv arxiv-ai | 1mo | On the Numerical Reliability of Differentiable Physics-Based Optimization for Robotic Material Manipulation | Differentiable physics is increasingly used in robotic material manipulation for system identification, trajectory or skill optimizati… | ↗Not briefed |
hn hn-ai | 1mo | LLMs in Professional Software Engineering | — | ↗Not briefed |
arxiv arxiv-ai | 1mo | Codoku: Renewable Program-Reasoning Challenges for Frontier Coding Agents | Existing program-reasoning benchmarks ask large language models to predict a program's behavior on a given input. Coding agents break… | ↗Not briefed |