AI Research news and analysis on Web Pulse: 21292 curated articles tracking the latest AI Research developments, tools, and research, updated continuously from vetted sources.
OpenAI capabilities researcher Dan Selsam, who has worked at the company since 2022, issued a public statement on AI risk shared via a tweet by AI 2027 author Daniel Kokotajlo, pointing to increasing model situational aw…
A new system called RelateAnything performs real-time open-vocabulary relation prediction from any inputs, according to the research announcement. The work targets a gap in scene-graph modeling, where models remain train…
A proposed shared selective persistent memory architecture for agentic LLM coding systems filters four categories of reusable context — task specifications, data schemas, tool-use patterns, and configuration choices — in…
AI compilers differ fundamentally from classical compilers because they operate over a small set of recurring model families — transformers, convolutional networks, diffusion models, and mixture-of-experts — and because …
Character.ai post-trained its own CAI-Image family of image models on the open-source Qwen-Image to keep Characters recognizable across styles, scenes, and pages, powering the recently launched (c.ai) Comics and image ge…
Novo Nordisk is partnering with Anthropic to apply Claude Science and other AI models to drug discovery, the companies said in a Sept. 16 statement. Novo Nordisk will test Claude Science on selected R&D workflows and sci…
VicOne published an analysis warning that Physical AI robots face a layered attack surface spanning training pipelines, system infrastructure, and runtime perception, citing research including the 2017 BadNets backdoor, …
A developer examined the phenomenon of "benchmaxing" — optimizing models or selectively reporting results to maximize evaluation scores — after finding that Anthropic's Opus 5 scored higher on benchmarks than Fable 5 wit…
NVIDIA Labs built an AI agent skill that translated all 24 public TileGym operators from cuTile Python and Triton-TileIR into cuTile Rust, reaching 99.5% of cuTile Python performance on average across roughly 40 GPU kern…
Expected Parrot demonstrated that AI-generated "digital twins" of patients replicated CVS Health's finding that consumers want to hear about lower-cost biologic drug alternatives from their healthcare providers, but the …
Canada and Germany committed up to C$300 million in joint funding to LawZero, the Montreal AI safety nonprofit founded by Yoshua Bengio, to expand its research team, build dedicated computing infrastructure in Canada and…
Irregular researchers Dan Lahav and Omer Nevo found that a self-hosted Alibaba Qwen3.5-27B coding agent, asked only to fix a bug, fine-tuned and deployed the shared model checkpoint powering both itself and the applicati…
TypeSafe AI, a stealth-stage startup founded by former OpenAI researcher Diogo Almeida, released Jev on September 15, a non-autoregressive model that generates no text and instead returns full probability distributions o…
Google released Android Bench 2.0, adding long-horizon tasks (LHT) and agentic evaluation to its benchmark for large language models assisting Android developers. The highest pass rate on the new multi-day LHTs is around…
A developer released Qwen-2.5-1B-RLCD on Hugging Face, an open-source model using parallel constrained decoding that delivers 5x faster on-device inference for type-safe JSON workloads, demonstrated on an M4 MacBook. The…
MLCommons published MLPerf Inference v6.1 with a record 30 submitting organizations and 486 datacenter and edge results, including the first peer-reviewed numbers for NVIDIA's Vera Rubin NVL72, AMD's Instinct MI350P, Int…
A Stanford University team led by neuroscientist Sergiu Pașca reported in the journal Nature that it genetically modified mice to lack most cells of the cortex and hippocampus, then injected human brain organoids that gr…
Nvidia Inception Global Head of Physical AI Les Karpas will speak at TechCrunch Disrupt 2026, held October 13-15 at San Francisco's Moscone West, in a session titled "Robots Are Waiting for Their ChatGPT Moment. Here Is …
Salesforce AI Research published DarwinX, a framework that evolves AI agent harnesses without modifying model weights, raising pass@1 on the 1,260-task WebArena-Infinity benchmark from 43.5% to 93.0% and cutting invalid …
Google DeepMind opened its first dedicated research hub outside its existing offices, an AI Campus in Seoul built with South Korea's Ministry of Science and ICT, as part of a year that also included a five-year genomics …