List your company

Common Sense Reasoning on WinoGrande (dev)

79.2AccuracyGLaM
54.947261.243667.5473.8364Dec 13, 2021Aug 31, 2022May 20, 2023Feb 5, 2024Oct 24, 2024Jul 12, 2025Mar 31, 2026

Evaluation results

MethodLinks
GLaMModel Size=64B/64E, Evaluation Protocol=Few-shot, Shots=162021.1279.2
Megatron-NLGModel Size=530B, Evaluation Protocol=Few-shot2021.1278.9
GPT-3Model Size=175B, Evaluation Protocol=Few-shot, Shots=162021.1277.7
GLaMModel Size=64B/64E, Evaluation Protocol=Zero-shot2021.1273.5
GPT-3Model Size=175B, Evaluation Protocol=One-shot2021.1273.2
GLaMModel Size=64B/64E, Evaluation Protocol=One-shot2021.1273.1
GPT-3Model Size=175B, Evaluation Protocol=Zero-shot2021.1270.2
GopherModel Size=280B, Evaluation Protocol=Few-shot, Shots=02021.1270.1
FusionGated-FIRNetDiscovery Source=AI Discovered2026.0360.14
Mamba2Discovery Source=Human Discovered2026.0358.48
Gated DeltaNetDiscovery Source=Human Discovered2026.0358.01
Content-SharpRouterDiscovery Source=AI Discovered2026.0357.85
AdaMulti-PathGateNetDiscovery Source=AI Discovered2026.0357.62
Hier-GateNetDiscovery Source=AI Discovered2026.0357.38
PathGate-FusionNetDiscovery Source=AI Discovered2026.0357.22
DeltaNetDiscovery Source=Human Discovered2026.0355.88