Jarrodbarnes/KernelBench-RLVR-120b

Partial fine-tune vs openai/gpt-oss-120b. Partial change: 66% of the weights changed (attention, lm_head), in 36 of 36 layers.

Tensors changed: MLP 0/144, Norms 0/73, Attention 135/324, lm_head 1/1, Embeddings 0/1

2 derivatives checked

• mradermacher/KernelBench-RLVR-120b-i1-GGUF: Quantization
• mradermacher/KernelBench-RLVR-120b-GGUF: Quantization

On the Hub · Downloads on Model Pulse