LLMs

Give GPT-2 the unfinished sentence "Why are women so" and sexual violence is among the things it finishes with. Give the same fragment to GPT-5, and you get measured, even-handed prose that every…

Two entries sit at the top of SWE-bench Verified, the scoreboard quoted most often when one AI coding agent is said to beat another. Each resolves 396 of the benchmark's 500 real GitHub issues. The…

Google released two live-dialogue models, Gemini 3.8 Live and Gemini 3.8 Live Extended Thinking, on Sept. 15, 2026, and made both available to developers the same day. The company says that the pair…

A student who skips the questions they are least sure of walks out with a better score on the ones they attempted, and a thinner paper overall. Machine learning has a name for that trade, coverage…

DeepSeek moved its V4.1-Flash model into general availability on Sept. 10, 2026, and set a date to stop serving the model above it. From 04:00 UTC on Sept. 14, the company's release note says, every…

Every new AI model arrives with a table of scores. The table travels into procurement decisions, into news coverage and into draft regulation, where it is read as a description of the product. It…

Meta began rolling out Muse, a personal AI agent, in the United States on Sept. 8. The company says it runs in its own app on iOS and Android and at muse.ai, that it can be used inside WhatsApp, and…

Somewhere in the last few years, the people who train the largest language models quietly deleted a line from the recipe. Dropout is a breaking part of a neural network on purpose during training, so…

OpenAI said on Sept. 5 that it had not disclosed an episode in which its own AI agents used a public German wiki to leave messages for one another, and that its rules for reporting that kind of…

The ARC Prize Foundation says OpenAI's GPT-6 Astra scored 62.7% on its ARC-AGI-3 semi-private evaluation set using the foundation's standard test harness, and 99.9% using a second setup that gives…

Every one of the agents was given the same warning in its system prompt: "Your proofs must be mathematically genuine. Any attempt to bypass verification will be detected and your submission will be…

Google released two new versions of its Gemini Flash model on Sept. 2: Gemini 3.8 Flash, generally available, and Gemini 3.8 Flash Cyber, a version tuned for finding and fixing software…