List your company

Scientific Reasoning on GPQA Diamond (Pass@1 accuracy)

43.43Pass@1 AccuracyTYLER Stage 2
14.018821.654429.2936.9256Jun 15, 2026

Evaluation results

MethodLinks
TYLER Stage 2Backbone=Qwen3-4B2026.0643.43
MemGenBackbone=Qwen3-4B2026.0640.4
GRPOBackbone=Qwen3-4B2026.0639.9
TYLER Stage 1Backbone=Qwen3-4B2026.0639.9
SoftCoTBackbone=Qwen3-4B2026.0638.38
TYLER Stage 2Backbone=SmolLM3-3B2026.0633.5
TYLER Stage 1Backbone=SmolLM3-3B2026.0630.5
GRPOBackbone=SmolLM3-3B2026.0628.28
MemGenBackbone=SmolLM3-3B2026.0625.25
SoftCoTBackbone=SmolLM3-3B2026.0624.24
SFTBackbone=SmolLM3-3B2026.0621.72
Soft-ThinkingBackbone=SmolLM3-3B2026.0621.72
SFTBackbone=Qwen3-4B2026.0621.72
SwiReasoningBackbone=SmolLM3-3B2026.0620.7
CoTBackbone=SmolLM3-3B2026.0619.7
SwiReasoningBackbone=Qwen3-4B2026.0617.17
CoTBackbone=Qwen3-4B2026.0616.16
Soft-ThinkingBackbone=Qwen3-4B2026.0615.15