List your company

Agent Memory Question Answering on PerLTQA (test)

42.68BLEUxMemory
10.148818.594427.0435.4856Feb 2, 2026

Evaluation results

MethodLinks
xMemoryModel=Llama-3.1-8B-Ins2026.0242.6852.3747.846,066.4
NemoriModel=Llama-3.1-8B-Ins2026.0241.0149.6244.6511,440.04
xMemoryModel=GPT-5 nano2026.0236.7946.2341.257,307.3
xMemoryModel=Qwen3-8B2026.0236.2447.0842.55,087.18
MemoryOSModel=Qwen3-8B2026.0235.1442.3538.486,499.47
MemoryOSModel=Llama-3.1-8B-Ins2026.0235.1442.3538.486,511.11
Naive RAGModel=Llama-3.1-8B-Ins2026.0233.6744.8439.659,531.3
NemoriModel=GPT-5 nano2026.0233.4441.7938.4311,882.7
A-MemModel=GPT-5 nano2026.0233.1241.1737.9414,717.77
NemoriModel=Qwen3-8B2026.0232.5542.838.059,091.72
Naive RAGModel=Qwen3-8B2026.0232.0841.3735.956,274.38
A-MemModel=Qwen3-8B2026.0228.6234.9933.549,864.32
MemoryOSModel=GPT-5 nano2026.0227.6633.8331.4412,669.39
Naive RAGModel=GPT-5 nano2026.0227.3537.7632.3510,756.39
LightMemModel=Llama-3.1-8B-Ins2026.0223.4735.9333.655,452.49
LightMemModel=Qwen3-8B2026.0216.9732.4129.877,692.05
LightMemModel=GPT-5 nano2026.0214.0428.326.117,579.27
A-MemModel=Llama-3.1-8B-Ins2026.0211.413.2815.97,707.4