List your company

Question Answering on HotpotQA (test) (Acc and R-L Metrics)

90.2AccuracyLATENTGRAPHMEM
11.57631.98852.472.812Jan 6, 2026

Evaluation results

MethodLinks
LATENTGRAPHMEMBackbone=Qwen3-8B2026.0190.271.21
LATENTGRAPHMEMBackbone=SmolLM3-3B2026.0188.961.39
PREMemBackbone=SmolLM3-3B2026.0187.760
LATENTGRAPHMEMBackbone=Qwen2.5-1.5B-Instruct2026.0186.658.54
PREMemBackbone=Qwen3-8B2026.0181.4656.53
Reasoner-onlyBackbone=Qwen2.5-1.5B-Instruct2026.0180.252.2
Reasoner-onlyBackbone=SmolLM3-3B2026.0177.151.46
Reasoner-onlyBackbone=Qwen3-8B2026.0172.662.48
MemGenBackbone=Qwen3-8B2026.0172.371.03
Mem0Backbone=Qwen2.5-1.5B-Instruct2026.0163.844.08
MemGenBackbone=SmolLM3-3B2026.015853.64
MemGenBackbone=Qwen2.5-1.5B-Instruct2026.0157.157.59
A-MemBackbone=SmolLM3-3B2026.0151.361.18
PREMemBackbone=Qwen2.5-1.5B-Instruct2026.0149.748.78
Mem0Backbone=Qwen3-8B2026.0148.629.43
A-MemBackbone=Qwen2.5-1.5B-Instruct2026.0146.656.73
Mem0Backbone=SmolLM3-3B2026.0144.929.3
THEANINEBackbone=SmolLM3-3B2026.0142.7919.93
THEANINEBackbone=Qwen3-8B2026.0139.9416.02
A-MemBackbone=Qwen3-8B2026.0131.455.1
RAGBackbone=Qwen3-8B2026.0129.363.88
THEANINEBackbone=Qwen2.5-1.5B-Instruct2026.0119.48.19
RAGBackbone=SmolLM3-3B2026.0115.352.22
RAGBackbone=Qwen2.5-1.5B-Instruct2026.0114.629.05