Upload SAE blocks.0.hook_attn_out
Browse files
blocks.0.hook_attn_out/cfg.json
CHANGED
@@ -1 +1 @@
|
|
1 |
-
{"architecture": "jumprelu", "d_in": 2048, "d_sae": 8192, "dtype": "bfloat16", "device": "cuda", "model_name": "meta-llama/Llama-3.2-1B-Instruct", "hook_name": "blocks.0.hook_attn_out", "hook_layer": 0, "hook_head_index": null, "activation_fn_str": "relu", "activation_fn_kwargs": {}, "apply_b_dec_to_input": false, "finetuning_scaling_factor": false, "sae_lens_training_version": "5.9.1", "prepend_bos": true, "dataset_path": "darpanaswal/TERInterpData", "dataset_trust_remote_code": true, "context_size": 128, "normalize_activations": "none", "neuronpedia_id": null, "model_from_pretrained_kwargs": {"center_writing_weights": false}, "seqpos_slice": [null]
|
|
|
1 |
+
{"architecture": "jumprelu", "d_in": 2048, "d_sae": 8192, "dtype": "bfloat16", "device": "cuda:1", "model_name": "meta-llama/Llama-3.2-1B-Instruct", "hook_name": "blocks.0.hook_attn_out", "hook_layer": 0, "hook_head_index": null, "activation_fn_str": "relu", "activation_fn_kwargs": {}, "apply_b_dec_to_input": false, "finetuning_scaling_factor": false, "sae_lens_training_version": "5.9.1", "prepend_bos": true, "dataset_path": "darpanaswal/TERInterpData", "dataset_trust_remote_code": true, "context_size": 128, "normalize_activations": "none", "neuronpedia_id": null, "model_from_pretrained_kwargs": {"center_writing_weights": false}, "seqpos_slice": [null]}
|
blocks.0.hook_attn_out/sae_weights.safetensors
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 67146120
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:362713bd3cd54656cc5b35ccfc1df7805da3c8c3f5e9a8042649ef0dbdcacb2d
|
3 |
size 67146120
|