List your company

Multi-hop Question Answering on LoCoMo Multi-Hop

48.15F1 ScoreMEMMA_LM
11.12620.73830.3539.962Mar 19, 2026

Evaluation results

MethodLinks
MEMMA_LMBackbone=GPT-4o-mini2026.0348.1539.6778.12
MEMMA_LMBackbone=Claude-Haiku-4.52026.0335.3832.4865.62
LightMemBackbone=Claude-Haiku-4.52026.0335.1131.8559.38
LightMemBackbone=GPT-4o-mini2026.0333.7429.3365.62
Full TextBackbone=GPT-4o-mini2026.0329.4121.1643.75
Full TextBackbone=Claude-Haiku-4.52026.0329.4121.1643.75
LangMemBackbone=Claude-Haiku-4.52026.0320.0514.8534.38
Naive RAGBackbone=GPT-4o-mini2026.0315.849.531.25
Naive RAGBackbone=Claude-Haiku-4.52026.0315.849.531.25
A-MemBackbone=Claude-Haiku-4.52026.0315.7910.3228.13
A-MemBackbone=GPT-4o-mini2026.0315.5610.8831.25
LangMemBackbone=GPT-4o-mini2026.0312.559.2225