-
-
-
-
-
-
Inference Providers
Active filters: topk_sae
r-takahashi/sae-baseline_llama_l4_d256_h2_seq1024-layer2-k32-latents32x-lr1e-3
4.2M • Updated
r-takahashi/sae-baseline_llama_l4_d512_h4_seq1024-layer2-k32-latents32x-lr1e-3
16.8M • Updated
r-takahashi/sae-baseline_llama_l6_d512_h4_seq1024-layer3-k32-latents32x-lr1e-3
16.8M • Updated
r-takahashi/sae-baseline_llama_l6_d512_h4_seq1024-layer4-k32-latents32x-lr1e-3
16.8M • Updated
r-takahashi/sae-baseline_llama_l6_d768_h6_seq1024-layer3-k32-latents32x-lr1e-3
37.8M • Updated
r-takahashi/sae-baseline_llama_l6_d768_h6_seq1024-layer4-k32-latents32x-lr1e-3
37.8M • Updated
r-takahashi/sae-baseline_llama_l8_d768_h6_seq1024-layer4-k32-latents32x-lr1e-3
37.8M • Updated
r-takahashi/sae-baseline_llama_l8_d768_h6_seq1024-layer6-k32-latents32x-lr1e-3
37.8M • Updated
r-takahashi/sae-llama_l8_d1024_h8_seq1024-layer4-k32-latents32x-lr1e-3
67.1M • Updated
r-takahashi/sae-llama_l8_d1024_h8_seq1024-layer6-k32-latents32x-lr1e-3
67.1M • Updated
r-takahashi/sae-llama_l8_d2048_h16_seq1024-layer4-k32-latents32x-lr1e-3
0.3B • Updated
r-takahashi/sae-llama_l8_d2048_h16_seq1024-layer6-k32-latents32x-lr1e-3
0.3B • Updated
r-takahashi/sae-llama_l16_d1024_h8_seq1024-layer8-k32-latents32x-lr1e-3
67.1M • Updated
r-takahashi/sae-llama_l16_d1024_h8_seq1024-layer14-k32-latents32x-lr1e-3
67.1M • Updated
r-takahashi/sae-llama_l16_d2048_h16_seq1024-layer8-k32-latents32x-lr1e-3
0.3B • Updated
r-takahashi/sae-llama_l16_d2048_h16_seq1024-layer14-k32-latents32x-lr1e-3
0.3B • Updated
r-takahashi/sae-llama_l24_d1024_h8_seq1024-layer12-k32-latents32x-lr1e-3
67.1M • Updated
r-takahashi/sae-llama_l24_d1024_h8_seq1024-layer22-k32-latents32x-lr1e-3
67.1M • Updated
r-takahashi/sae-llama_l24_d2048_h16_seq1024-layer12-k32-latents32x-lr1e-3
0.3B • Updated
r-takahashi/sae-llama_l24_d2048_h16_seq1024-layer22-k32-latents32x-lr1e-3
0.3B • Updated
r-takahashi/sae-speedrun_d6_chatsft-layer4-k32-latents32x-lr1e-3
9.45M • Updated
r-takahashi/sae-speedrun_d6_chatsft-layer1-k32-latents32x-lr1e-3
9.45M • Updated
r-takahashi/sae-speedrun_d6_chatsft-layer2-k32-latents32x-lr1e-3
9.45M • Updated
r-takahashi/sae-speedrun_d6_chatsft-layer3-k32-latents32x-lr1e-3
9.45M • Updated
r-takahashi/sae-speedrun_d6_chatsft-layer5-k32-latents32x-lr1e-3
9.45M • Updated
r-takahashi/sae-speedrun_d6_chatsft-layer6-k32-latents32x-lr1e-3
9.45M • Updated
r-takahashi/sae-speedrun_d6_chatsft-layer1-k32-latents65536-lr1e-3
50.4M • Updated
r-takahashi/sae-speedrun_d6_chatsft-layer2-k32-latents65536-lr1e-3
50.4M • Updated
r-takahashi/sae-speedrun_d6_chatsft-layer3-k32-latents65536-lr1e-3
50.4M • Updated
r-takahashi/sae-speedrun_d6_chatsft-layer4-k32-latents65536-lr1e-3
50.4M • Updated