List your company

Multi-session Conversation Question Answering on LoCoMo 98 QA probes

2.18Score (1-5)Bookmark+Recall
1.61841.76421.912.0558Apr 14, 2026

Evaluation results

MethodLinks
Bookmark+RecallLLM=GPT-4o-mini2026.042.181.89
Full ContextContext Window=truncated to 60 turns, LLM=GPT-4o-mini2026.042.021.75
Search-Tool BaselineLLM=GPT-4o-mini2026.041.91.61
Word-Overlap RetrievalRetrieval=top-3, LLM=GPT-4o-mini2026.041.881.63
BM25 RetrievalRetrieval=top-3, LLM=GPT-4o-mini2026.041.861.61
TruncationContext Window=20 turns, LLM=GPT-4o-mini2026.041.641.43