models
Open weights, fine-tunes and adapters. Every listing here comes live from the Hugging Face Hub, attributed to it, and links back to the source.
Qwen3-14B-Per-Layer-Refusal-Directions-alpha-1llama2_7b_perlayer_tucker500uce-4-layerqwen3.5-0.8B-heretic-per-layerBRP-Sochirca-CodeGPT-Py150-0.6-sparse-q-all-layers-asym-per-channelBRP-Sochirca-CodeGPT-Py150-0.6-sparse-q-all-layers-asym-per-tensorBRP-Sochirca-CodeGPT-Py150-0.6-sparse-q-all-layers-sym-per-channelBRP-Sochirca-CodeGPT-Py150-0.6-sparse-q-all-layers-sym-per-tensorCodeGPT-Py150_q_all_layers_sym_per_tensoruce-33-layeradapt-distill-bridge-perlayer-460m-5000lingbot-40-per-layer-x4-15000LAT_4-3_sweep_1_pgd_layers_0_epsilon_0.4_pgd_iterations_per_step_2LAT_4-3_sweep_1_pgd_layers_25_epsilon_0.2_pgd_iterations_per_step_28adapt-distill-bridge-perlayer-460m-2500LAT_4-3_sweep_1_pgd_layers_0_6_12_18_24_epsilon_0.05_pgd_iterations_per_step_16Mixtral-8x7B_Instruct-v0.1_pruning_per_layer_4expertsqwen-finance-per-layer-top-t-32-key-value-num-tokens-512LAT_4-3_sweep_1_pgd_layers_18_epsilon_0.5_pgd_iterations_per_step_22Mixtral-8x7B_Instruct-v0.1_pruning_per_layer_7expertsMixtral-8x7B_Instruct-v0.1_pruning_per_layer_6expertsadapt-distill-bridge-perlayer-460m-1500LAT_4-3_sweep_1_pgd_layers_8_epsilon_0.5_pgd_iterations_per_step_10LAT_4-3_sweep_1_pgd_layers_31_epsilon_25.0_pgd_iterations_per_step_22LAT_4-3_sweep_1_pgd_layers_25_epsilon_0.2_pgd_iterations_per_step_2LAT_4-3_sweep_1_pgd_layers_25_epsilon_0.4_pgd_iterations_per_step_22LAT_4-3_sweep_1_pgd_layers_29_epsilon_0.7_pgd_iterations_per_step_28LAT_4-3_sweep_1_pgd_layers_25_epsilon_0.7_pgd_iterations_per_step_16lingbot-20-per-layer-x4-15000adapt-distill-bridge-perlayer-460m-1500-student