1
OckBench: Measuring the Efficiency of LLM Reasoning
关注大模型推理的token效率，填补了仅重准确率、忽略效率的基准空白。
arXiv:2511.05722v3 Announce Type: replace-cross Abstract: Large language models (LLMs) such as GPT-5 and Gemini 3 have pushed the frontier of automate…