List your company

STEM Reasoning on SuperGPQA (Pass@1)

32Pass@1 AccuracyTAC
14.3218.9123.528.09Jun 23, 2026

Evaluation results

MethodLinks
TACBackbone=Qwen3-4B-Base, Seeds=12026.0632
RandomBackbone=Qwen3-4B-Base, Seeds=12026.0630.5
SECBackbone=Qwen3-4B-Base, Seeds=12026.0629.9
TACBackbone=Qwen3-1.7B2026.0622.1
RandomBackbone=Qwen3-1.7B2026.0621.9
SECBackbone=Qwen3-1.7B2026.0621.9
M2OBackbone=Qwen3-1.7B2026.0621.7
M2OBackbone=Llama3.2-3B2026.0621.5
TACBackbone=Llama3.2-3B2026.0621.3
BaseBackbone=Llama3.2-3B2026.0621
RandomBackbone=Llama3.2-3B2026.0618.8
BaseBackbone=Qwen3-1.7B2026.0618.5
SECBackbone=Llama3.2-3B2026.0618
TACBackbone=Qwen3-0.6B-Base, Seeds=12026.0616.9
RandomBackbone=Qwen3-0.6B-Base, Seeds=12026.0615.6
SECBackbone=Qwen3-0.6B-Base, Seeds=12026.0615