New Benchmark EviCover Targets Evidence Loss in Long-Term Conversational Memory
Researchers introduce EviCover, a benchmark for long-term conversational question answering where several separated dialogue turns must jointly support an answer, diagnosing 'proof-set loss' in long-term conversational memory.
Read the full story at arXiv cs.CL ↗
Timeline · 1 report
- 2026-11-10 05:00 · arXiv cs.CL
Where Did the Evidence Go? Diagnosing Proof-Set Loss in Long-Term Conversational Memory