往下拉回到首頁
有人直接動刀改造 Qwen AI 大腦,移除限制後的版本現在在網上瘋傳

有人直接動刀改造 Qwen AI 大腦,移除限制後的版本現在在網上瘋傳

Qwen3.6-35B-A3B-Plus-Uncensored-Wasserstein (neuron level surgery)

Hello everyone. During data debugging session on per tensor and per neuron level I found that neurons in tensor layers in MoE model can die (have zero value). Here the log. For example In blk.0.ffn_gate_exps.weight and blk.0.ffn_up_exps.weight in Qwen3.6 35B A3B Q8_0 quant: I found 40% of zero neurons. In Qwen3.5 9B I didn't found any zero blocks. All blocks in it contain value. Don't know why this is happened. I never trained LLM's by myself, but this problem exists. A company I'm intervie