Hacker News new | ask | show | jobs
by mrciffa 498 days ago
Yes, reasoning models can potentially be optimized with our uncertainty estimations. We are currently testing the library with DeepSeek R1