Share your thoughts, 1 month free Claude Pro on usSee more
WorkDL logo mark

Long-context Memory Retrieval and Reasoning on LongMemEval 128K

47.26F1 Score

MemSifter

30.994435.217239.4443.6628Mar 3, 2026
Updated 1mo ago

Evaluation Results

MethodLinks
2026.03
47.26
2026.03
46.09
2026.03
44.26
2026.03
43.04
2026.03
42.2
2026.03
41.75
2026.03
40.46
2026.03
40.13
2026.03
39.55
2026.03
38.56
2026.03
35.38
2026.03
35.15
2026.03
34.39
2026.03
34.31
2026.03
33.93
2026.03
33.64
2026.03
33.53
2026.03
32.62
2026.03
31.96
2026.03
31.62