List your company

Cross-Task Evaluation on Alpaca, GSM8K, and MBPP

49.1Average AccuracyOriginal Merged Model
0.386413.033225.6838.3268Jun 29, 2026

Evaluation results

MethodLinks
Original Merged ModelBackbone=Qwen2-7B, Protect=-, Attack=-2026.0649.1—
AGABackbone=Qwen2-7B, Protect=MergeLock [36], Attack=AGA2026.0649.0146.75
AGABackbone=Qwen2-7B, Protect=Params [18], Attack=AGA2026.0648.9345.85
AGABackbone=Qwen2-7B, Protect=Params-D [18], Attack=AGA2026.0647.7944.72
AGABackbone=Qwen2-7B, Protect=MergeBarrier [22], Attack=AGA2026.0647.0931.29
MergeBarrierBackbone=Qwen2-7B, Protect=MergeBarrier [22], Attack=-2026.0615.8—
ParamsBackbone=Qwen2-7B, Protect=Params [18], Attack=-2026.063.08—
Params-DBackbone=Qwen2-7B, Protect=Params-D [18], Attack=-2026.063.07—
MergeLockBackbone=Qwen2-7B, Protect=MergeLock [36], Attack=-2026.062.26—