FACTS Benchmark Suite: Systematically evaluating the factuality of large language models
Systematically evaluating the factuality of large language models with the FACTS Benchmark Suite.
en
LLMs
3025
249
How confessions can keep language models honest
OpenAI researchers are testing “confessions,” a method that trains models to admit when they make mistakes or act undesirably, helping improve AI honesty, transparency, and trust in model outputs.
en
Introducing AutoJudge: Streamlined inference acceleration via automated dataset curation
AutoJudge accelerates LLM inference by identifying which token mismatches actually matter. Using self-supervised learning to train a lightweight classifier, it accepts up to 40 draft tokens per cycle—delivering 1.5–2× speedups over standard speculative decoding with minimal accur
en
How to evaluate and benchmark Large Language Models (LLMs)
Understanding how to evaluate and benchmark Large Language Models (LLMS). Test, compare, and understand LLMs.
en
RL without TD learning
In this post, I’ll introduce a reinforcement learning (RL) algorithm based on an “alternative” paradigm: divide and conquer . Unlike traditional methods, this algorithm is not based on temporal difference (TD) learning (which has scalability challenges ), and scales well to long-horizon tasks. We can do Reinforcement Learning (RL) based on divide and conquer, instead of temporal difference (TD) learning.…
en
Dynamic AI agent testing for the real world with Collinear Simulations and Together Evals
Test AI agents in the real world with Collinear TraitMix and Together Evals: dynamic persona simulations, multi-turn dialogs, and LLM-as-judge scoring.
en
Why language models hallucinate
OpenAI’s new research explains why language models hallucinate. The findings show how improved evaluations can enhance AI reliability, honesty, and safety.
en
What exactly does word2vec learn?
What exactly does word2vec learn, and how? Answering this question amounts to understanding representation learning in a minimal yet interesting language modeling task. Despite the fact that word2vec is a well-known precursor to modern language models, for many years, researchers lacked a quantitative and predictive theory describing its learning process. In our new paper , we finally provide such a…
en
Introducing gpt-oss
We’re releasing gpt-oss-120b and gpt-oss-20b—two state-of-the-art open-weight language models that deliver strong real-world performance at low cost. Available under the flexible Apache 2.0 license, these models outperform similarly sized open models on reasoning tasks, demonstrate strong tool use capabilities, and are optimized for efficient deployment on consumer hardware.
en
Toward understanding and preventing misalignment generalization
We study how training on incorrect responses can cause broader misalignment in language models and identify an internal feature driving this behavior—one that can be reversed with minimal fine-tuning.
en