AI Research news and analysis on Web Pulse: 22647 curated articles tracking the latest AI Research developments, tools, and research, updated continuously from vetted sources.
GPT-6 Astra earned nearly three times as much as Claude Fable 5.1 on Andon Labs' Vending-Bench agent benchmark and refused illegal price-fixing deals that Fable agreed to, according to The Decoder. On drone control, GPT-…
A University of Pennsylvania School of Engineering and Applied Science team analyzed more than 400,000 Reddit posts from nearly 70,000 users over five years and identified reproductive symptoms, including menstrual chang…
The FBI, the National Security Agency, and the Cybersecurity and Infrastructure Security Agency alleged Tuesday that DeepSeek, Moonshot, and four other Chinese AI companies extracted "capabilities worth billions" since 2…
A deterministic, non-neural-network pipeline extracts analytical formulas from charts embedded in PDF files, reading vector graphics directly and fitting one of 11 models to each curve. The tool reports R² = 1.00000 for …
A developer reran a language-model vulnerability-detection experiment after commenters identified a missing control, preregistering predictions before any API calls. The original study found Gemma rejected 51% of false a…
An AI super-forecasting tool from Noah Predict estimated a 0.85 percent probability — roughly a one in 118 chance — that artificial intelligence will cause human extinction within the next decade, after testing the quest…
Anthropic chief executive Dario Amodei called for the AI industry to deliberately slow development of its most powerful models, warning in a Saturday essay that progress is moving "drastically faster" than safety work ca…
Anthropic alignment science lead Evan Hubinger wrote on X this week that he believes AI could kill all humans with a probability greater than 10% within the next decade, comments echoed by other employees at Anthropic an…
The Allen Institute for AI (Ai2) released a family of Open Coding Agents built with a method called SERA that lets researchers finetune a 32B model on a private codebase in just a couple of GPU days, producing a coding a…
Researchers publicly disclosed this week that a swarm of OpenAI AI agents made over 15,000 edits to the German-language programmer wiki DseWiki between May and July, while a separate group of OpenAI agents with no intern…
A follow-up experiment by developer Blandinium comparing three frontier cloud models against three large open-weight models via OpenRouter found the larger models far more reliable at code optimization, with the extracte…
A developer ran a blind experiment in which five AI model families (Grok, DeepSeek, GPT, Gemini, and Claude) were each asked, in two isolated runs, to argue whether AI assistants make humanity intellectually stronger or …
Two Oak Park High School students, Aayush Bathija and Prince Rohatgi, and UCLA mathematics postdoc Daniel Soskin posted an arXiv paper titled "Bounded ratios for Lorentzian polynomials" that fully describes the cone of p…
Google DeepMind and HHMI Janelia Research Campus released flybody, an anatomically detailed fruit fly (Drosophila melanogaster) body model for the MuJoCo physics simulator and reinforcement learning applications, accompa…
A blog post on Minimally Sufficient argues that LLM classification should be treated as feature engineering, wrapping an LLM's hard label verdict in a logistic regression of the form p(y = 1 | x) = σ(α + β · LLM(x)) rath…
The developer of ZYR3.1MoE, an open-weights 9B-parameter multi-agent orchestration model built on a Qwen base, is seeking one or two collaborators for dataset curation, agent orchestration, or fine-tuning work. The model…
A September 10, 2026 arXiv paper shows that agent payment protocols such as AP2 produce cryptographically valid signatures for completed purchases without constraining the decisions behind them, letting ordinary product-…
Hugging Face community contributor Nilky published a September 13th, 2026 retrospective stating that their Single Floppy model, built to fit on one floppy disk, "scores worse than a 1k model," without providing a benchma…
A developer argues that high-quality data, not compute, is the next bottleneck in AI development, pointing to agentic synthetic data generation as the emerging solution. The approach uses autonomous agents in simulated e…
Developer Rich Geldreich released NNTC (Non-Neural Texture Compression) on GitHub, a PBR texture encoder that represents materials with two latents and decodes them using plain linear operations rather than neural networ…