List your company

Question Answering on LongMemEval S (test)

84.21QA Score (TR Context)LightMem
29.474843.684957.89572.1051May 15, 2026

Evaluation results

MethodLinks
LightMemBackbone=GPT-4.1-mini2026.0584.2151.1388.469023.2176.6769.6
SimpleMemBackbone=GPT-4.1-mini2026.0581.9557.8974.3682.8671.4376.6773
DA-UpdateBackbone=GPT-4.1-mini2026.0577.4469.1788.4698.5773.2156.6778.2
DimMemBackbone=Qwen3-30B-A3B-Instruct-25072026.0576.6954.8979.4994.2973.2156.6772.2
DimMem-4B†Extractor=Qwen3-4B, QA Model=Qwen3-4B2026.0576.6955.6483.3384.2942.8656.6768.2
LightMemBackbone=Qwen3-30B-A3B-Instruct-25072026.0575.9457.8989.7485.7121.4373.3368.4
DimMemBackbone=GPT-4.1-mini2026.0574.4464.6684.6298.5776.7963.3376.4
SimpleMemBackbone=Qwen3-30B-A3B-Instruct-25072026.0574.4452.6375.6487.1464.2966.6769
DimMem-4BExtractor=Qwen3-4B, QA Model=GPT-4.1-mini2026.0573.6867.4283.339046.436071.8
MemOSBackbone=Qwen3-30B-A3B-Instruct-25072026.0572.9351.1373.088026.7953.3361.8
A-MemBackbone=Qwen3-30B-A3B-Instruct-25072026.0566.9244.3671.7974.2935.7163.3359
NaiveRAGBackbone=Qwen3-30B-A3B-Instruct-25072026.0565.4148.8771.7981.4330.3653.3359.6
MemOSBackbone=GPT-4.1-mini2026.0563.9154.1470.519064.295065.2
FullTextBackbone=Qwen3-30B-A3B-Instruct-25072026.0563.9142.1165.3871.4332.1446.6754.8
A-MemBackbone=GPT-4.1-mini2026.0558.6552.6369.2388.5757.1446.6762
NaiveRAGBackbone=GPT-4.1-mini2026.0557.1454.1474.3677.1460.7153.3362
MemoryOSBackbone=Qwen3-30B-A3B-Instruct-25072026.0557.1440.664.161.4328.575050.8
MemoryOSBackbone=GPT-4.1-mini2026.0550.3848.1271.7985.7151.7966.6759.2
FullTextBackbone=GPT-4.1-mini2026.0531.5846.6243.598041.0756.6746.8