论文精选01:48MerchantBench:评测电商LLM智能体长期连贯性的新基准MerchantBench 专门测 LLM 智能体在电商里的长期连贯性,比单轮问答更贴近实际运营,可以看看。#MerchantBench#LLM#智能体#电商AK@_akhaliq原文稍后读已读值得跟进有用关注 MerchantBench