Models09:32Chinese contentRUMBA:俄语用户记忆基准,细粒度评估长对话记忆想测LLM的长对话记忆力?RUMBA这个俄语基准能细粒度诊断模型在时间推理上的强项和短板。#RUMBA#俄语基准#长对话记忆#推理aarXiv cs.AI@Elizaveta Shevtsova et al. (5)SourceRead laterReadWorth followingUsefulFollow RUMBA