Advancing LLM Reasoning Generalists with Preference Trees

arXiv:2404.02078v1 Announce Type: cross
Abstract: We introduce Eurus, a suite of large language models (LLMs) optimized for reasoning. Finetuned from Mistral-7B and CodeLlama-70B, Eurus models achieve state-of-the-art results among open-source models on a diverse set of benchmarks covering mathematics, code generation, and logical reasoning problems. Notably, Eurus-70B beats GPT-3.5 Turbo in reasoning through a comprehensive benchmarking across 12 tests covering five tasks, and achieves a 33.3% pass@1 accuracy on LeetCode and 32.6% on TheoremQA, two challenging benchmarks, substantially outperforming existing …

70b abstract art arxiv benchmarks code code generation codellama cs.ai cs.cl cs.lg diverse gpt gpt-3 gpt-3.5 language language models large language large language models llm llm reasoning llms mathematics mistral open-source models reasoning results set state through trees turbo type