List your company

Student Simulation on FalconCode 1.0 (test)

98.2CoverageDPO
31.43248.76666.183.434Apr 12, 2026

Evaluation results

MethodLinks
DPOModel=Qwen3-4B2026.0498.279.770-0.024-0.079-0.137
DPOModel=Qwen3-8B2026.0495.779.671.4-0.061-0.074-0.129
SFTModel=Qwen3-4B2026.0492.278.270.4-0.107-0.06-0.138
PARAModel=Qwen3-8B2026.0492.178.571.4-0.112-0.059-0.121
SFTModel=Qwen3-8B2026.0492.179.371.8-0.11-0.062-0.126
DAPOModel=Qwen3-8B2026.0491.979.670.9-0.116-0.078-0.133
PARAModel=Qwen3-4B2026.0491.876.369.5-0.118-0.071-0.137
DAPOModel=Qwen3-4B2026.0485.978.170-0.203-0.068-0.138
BASEModel=Qwen3-8B2026.045773.358.1-0.5810.287-0.046
BASEModel=Qwen3-4B2026.0452.86949.1-0.640.278-0.063
BASEModel=GPT-5-mini2026.043454.943.3-0.910.184-0.078