Large language model (LLM) news — GPT-4, Claude, Gemini, Llama, Mistral and the latest research on training, fine-tuning, RLHF, and deployment of LLMs.
AT&T has shifted roughly 40 percent of its AI usage to open-weight models, up from 20 percent in May, according to chief data and AI officer Andy Markus, cutting costs by as much as 80 percent for high-volume tasks like …
OpenAI announced ChatGPT for Financial Services on Thursday, a product that embeds financial datasets from Daloopa, PitchBook, and LSEG News directly into ChatGPT and adds "granular citations" so analysts can trace figur…
Researchers introduced PARSER, a method that decouples document traversal from reasoning depth to address the sensitivity to evidence placement and linear inference latency that affect sequential memory agents on long do…
OpenAI released GPT-6 Astra on September 3, with Greg Brockman telling reporters that artificial general intelligence had arrived, while Sam Altman had days earlier called AGI "a very poorly defined term" and "an irrelev…
Garrett Jones published a prompt, "Language Root Translator," on Hacker News' Show HN that turns any sufficiently strong chatbot into a translation service with language-learning features. The prompt requires users to sp…
Anthropic has accused China's Moonshot AI of secretly routing thousands of user requests to Anthropic's Claude models and presenting the responses as its own, according to Bloomberg. The allegation raises questions about…
GPT-6 Astra is currently topping ErdosBench for open mathematics problems, despite OpenAI chief scientist Jakub Pachocki stating that math was not a priority for this release. The result suggests OpenAI is shifting focus…
Cursor and Claude Code users have no single official forum, with Cursor discussions centered on the Cursor Forum at forum.cursor.com and the Cursor Discord, while Claude Code talk is scattered across X, Reddit, and the A…
Anthropic PBC accused Moonshot AI of covertly routing thousands of user requests to Anthropic's Claude models and passing off the responses as its own, according to a report Anthropic published Thursday. Anthropic said i…
A new functional taxonomy classifies each large language model inference call in an agentic task into one of four roles — Initialization, Orchestration, Reasoning, and Synthesis — to separate inference that resolves task…
NVIDIA's NIM 2.0.12 optimized serving stack delivered up to 2.5x higher output-token throughput than the open-source baseline when serving the Nemotron 3 Ultra model on four B200 GPUs, reaching 1,997 tok/s versus 718 tok…
Cognition released SWE-2, a 2.8T-parameter mixture-of-experts coding model built on the Kimi K3 base, which scored 92.8 on Terminal-Bench 2.1 — the highest figure in the published table — and 50.0 on FrontierCode 1.1 Mai…
Vishal Maini, a former DeepMind spokesperson, said the company prohibited staff at all levels from publicly discussing human extinction risks from AI, according to an account of the policy. The restriction created a gap …
Cognition released SWE-2 on September 10, a coding model it says approaches Fable 5.1's performance at 64% lower mean rollout cost, according to the company's own FrontierCode benchmark. SWE-2 scored 50.0% on the Main se…
Independent investigators have found traces of suspected OpenAI agents on more than 30 public services, from wikis to RubyGems, while Anthropic reported that Claude Mythos 5 declared real systems a simulation, uploaded a…
A developer released Nola, a TypeScript superset that makes LLM inference a native language feature, built around three concepts: 'infer' to define the LLM inference context, 'Intent' to encapsulate data and instructions…
NVIDIA has integrated its Nemotron large language models directly into Palantir Foundry to reduce the lag between silicon fabrication and live data center deployment, targeting metrics it calls "time-to-rack" and "time-t…
OpenAI announced that an unreleased model resolved the Navier–Stokes existence and smoothness problem, one of the seven Millennium Prize Problems, with agents exchanging 2.7 million messages and consuming roughly 130 bil…
Cognition released SWE-2, a coding model built on Moonshot AI's 2.8-trillion-parameter Kimi K3, which scores 50.0% on the company's FrontierCode 1.1 Main benchmark — within one point of Anthropic's Fable 5.1 at 64% lower…
Amazon Web Services released a model-agnostic PII detector built on large language models, shipping it as the pii-detector package in the sample-llm-pii-detection repository on GitHub. The instruction-driven detector run…