On this day: May 22

The first follow-up since Anthropic's April 2025 disclosure of unfaithfulness to propose interventions at the training stage

The first case of a voluntary framework constraining the conditions of an actual deployment

Company-reported: SWE-bench Verified 72.5% for Opus 4 and 72.7% for Sonnet 4, Terminal-bench 43.2% for Opus 4. Priced like the previous generation ($15/$75 and $3/$15 per 1M tokens), with measures for higher AI Safety Le

A compromise between MHA and MQA that allows uptraining of existing checkpoints, and became the industry-standard KV-saving technique

Restructured the RNN so it can be trained at Transformer level; the opening signal of a revival of linear complexity

The canonical text of retrieval-augmented generation, combining parametric and non-parametric memory