Nathan Lambert

This talk covers the crucial details it takes to train a 7B parameter, fully open reasoning model, Olmo-Thinking, to rival Qwen 3, highlighting fresh results, trade-offs, and methods across midtraining, distillation with high-quality thinking SFT data, and reinforcement learning with verifiable...

2025 has represented an inflection point year for open models — spearheaded by the release of DeepSeek R1 in January, but built upon by a rapid expansion of the Chinese open model ecosystem. The rapid increase in capabilities is largely driven by a handful of Chinese companies in addition to...

Current AI models are extremely skilled, which was seen as the step change in evaluation scores across the industry in the first half of 2025, but often fail when presented with even medium time-horizon tasks. This talk presents a taxonomy of 4 traits of reasoning models -- skills, calibration, strategy, and abstraction -- that will be crucial to creating the next generation of AI applications.