Reinforcement learning

Hallucinations persist because benchmarks reward confident guessing

Why do LLMs keep getting things confidently wrong when saying "I don't know" would serve everyone better? Researchers at OpenAI offer a clear answer: the root of the problem is statistical. It appears during pretraining and is then locked in by the way we evaluate models after fine-tuning. In short: the data can be free of errors and the training objective will still push the model toward…

2026-09-19

Retrieve-for-Train compiles RL rewards into a 53.9M retriever

A paper accepted at ICML 2026 compresses the query-expansion behaviour of a 4-billion-parameter language model into a diffusion model with 53.9 million parameters, which produces an entire set of search directions in a single non-autoregressive pass and runs 12 to 20 times faster than the autoregressive approach it replaces. The method is called Retrieve-for-Train, and the paper is "Efficient,…

2026-09-19