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

Long-context Reasoning on ∞ Bench

90.39Accuracy

MiA (Emb-Only)

Updated 1mo ago

Evaluation Results

Method	Links
MiA (Emb-Only) 2025.12		90.39
MiA 2025.12		88.21
MiA (Emb-Only) 2025.12		87.77
Vanilla 2025.12		86.9
MiA (Gen-Only) 2025.12		86.46
MiA 2025.12		86.46
MiA-RAG 2025.12		85.15
MiA (Emb-Only) 2025.12		84.72
MiA 2025.12		84.71
MiA (Emb-Only) 2025.12		84.28
MiA 2025.12		84.28
MiA (Gen-Only) 2025.12		83.84
MiA 2025.12		83.41
MiA (Emb-Only) 2025.12		82.97
MiA (Gen-Only) 2025.12		82.1
MiA (Gen-Only) 2025.12		82.1
MiA-RAG 2025.12		81.22
Vanilla 2025.12		80.79
MiA (Gen-Only) 2025.12		80.79
MiA (Gen-Only) 2025.12		80.79
MiA-RAG 2025.12		80.79
MiA 2025.12		79.91
MiA (Gen-Only) 2025.12		79.48
MiA (Emb-Only) 2025.12		79.48
Vanilla 2025.12		77.29
MiA (Gen-Only) 2025.12		75.98
MiA (Gen-Only) 2025.12		75.98
Vanilla 2025.12		75.55
Vanilla 2025.12		73.8
Vanilla 2025.12		72.49
Summary-Only 2025.12		72.05
Summary-Only 2025.12		61.57