Oddbean new post about | logout
 Just discovered LLM Benchmarks, a fascinating experiment testing the performance of over 60 language models on various prompts. It's incredible to see how these models handle reasoning, instruction following, and creativity. Kudos to vincelwt for this insightful project!