List your company

Instruction Following on IFEval MultiIF

90.65I-Following ScoreA-SFT → A-GRPO
55.258864.446973.63582.8231Sep 29, 2025

Evaluation results

MethodLinks
A-SFT → A-GRPOBackbone=Qwen3-4B2025.0990.65
SFT → A-GRPOBackbone=Qwen3-4B2025.0990.45
SFT → GRPOBackbone=Qwen3-4B2025.0989.75
A-CHORDBackbone=Qwen3-4B2025.0988.27
A-GRPOBackbone=Qwen3-4B2025.0988.17
CHORDBackbone=Qwen3-4B2025.0986.97
GRPOBackbone=Qwen3-4B2025.0986.9
A-SFTBackbone=Qwen3-4B2025.0982.47
SFTBackbone=Qwen3-4B2025.0980.6
Qwen3-4BBackbone=Qwen3-4B2025.0980.1
A-SFT → A-GRPOBackbone=Qwen3-0.6B2025.0979.86
A-GRPOBackbone=Qwen3-0.6B2025.0979.58
SFT → A-GRPOBackbone=Qwen3-0.6B2025.0979.57
SFT → GRPOBackbone=Qwen3-0.6B2025.0977.4
A-CHORDBackbone=Qwen3-0.6B2025.0975.25
GRPOBackbone=Qwen3-0.6B2025.0974.09
CHORDBackbone=Qwen3-0.6B2025.0973.52
A-SFTBackbone=Qwen3-0.6B2025.0960.72
SFTBackbone=Qwen3-0.6B2025.0959.97
Qwen3-0.6BBackbone=Qwen3-0.6B2025.0956.62