List your company

Commonsense Reasoning on COMM

43.73AccuracyVERSATILEFFN
32.321235.283138.24541.2069Dec 16, 2025

Evaluation results

MethodLinks
VERSATILEFFNBase Model Params=1.21B, Protocol=Zero-shot2025.1243.73
4-LOOPBase Model Params=1.21B, Protocol=Zero-shot2025.1242.86
6-LOOPBase Model Params=1.21B, Protocol=Zero-shot2025.1241.93
MOEBase Model Params=1.21B, Protocol=Zero-shot2025.1241.2
VERSATILEFFNBase Model Params=720.81M, Protocol=Zero-shot2025.1240.79
2-LOOPBase Model Params=1.21B, Protocol=Zero-shot2025.1240.7
BASEBase Model Params=1.21B, Protocol=Zero-shot2025.1240.05
4-LOOPBase Model Params=720.81M, Protocol=Zero-shot2025.1239.39
6-LOOPBase Model Params=720.81M, Protocol=Zero-shot2025.1238.98
2-LOOPBase Model Params=720.81M, Protocol=Zero-shot2025.1237.84
MOEBase Model Params=720.81M, Protocol=Zero-shot2025.1237.1
VERSATILEFFNBase Model Params=354.71M, Protocol=Zero-shot2025.1236.69
4-LOOPBase Model Params=354.71M, Protocol=Zero-shot2025.1236.53
6-LOOPBase Model Params=354.71M, Protocol=Zero-shot2025.1235.95
2-LOOPBase Model Params=354.71M, Protocol=Zero-shot2025.1235.46
BASEBase Model Params=720.81M, Protocol=Zero-shot2025.1235.14
MOEBase Model Params=354.71M, Protocol=Zero-shot2025.1234.64
BASEBase Model Params=354.71M, Protocol=Zero-shot2025.1232.76