List your company

Multi-hop Question Answering on HotpotQA 800 documents

55.46F1 ScoreE-mem
21.805630.542839.2848.0172Jan 29, 2026

Evaluation results

MethodLinks
E-memBackbone=GPT4o-mini2026.0155.46
GAMBackbone=GPT4o-mini2026.0152.86
RAGBackbone=GPT4o-mini2026.0151.84
Long-ContextBackbone=GPT4o-mini2026.0149.71
E-memBackbone=Qwen2.5-14B2026.0147.91
Long-ContextBackbone=Qwen2.5-14B2026.0146.82
RAGBackbone=Qwen2.5-14B2026.0146.72
GAMBackbone=Qwen2.5-14B2026.0141.1
LIGHTMEMBackbone=GPT4o-mini2026.0135.28
Mem0Backbone=Qwen2.5-14B2026.0132.44
Mem0Backbone=GPT4o-mini2026.0131.74
MEMORYOSBackbone=Qwen2.5-14B2026.0130.25
A-memBackbone=GPT4o-mini2026.0130.2
LIGHTMEMBackbone=Qwen2.5-14B2026.0127.72
A-memBackbone=Qwen2.5-14B2026.0125.65
MEMORYOSBackbone=GPT4o-mini2026.0123.1