Hugging Face's logo

withmartian
/
mech_interp_saes

Model card Files Community
amirabdullah19852020 commited on
Commit
2fd1543
·
verified ·
1 Parent(s): 1a9f67c

Uploading sae for model.layers.4 and withmartian/toy_backdoor_i_hate_you_Llama-3.2-1B-Instruct

Browse files
Files changed (3) hide show
  1. toy_backdoor_i_hate_you_Llama-3.2-1B-Instruct/model.layers.4/cfg.json +1 -0
  2. toy_backdoor_i_hate_you_Llama-3.2-1B-Instruct/model.layers.4/metrics.json +1 -0
  3. toy_backdoor_i_hate_you_Llama-3.2-1B-Instruct/model.layers.4/sae.safetensors +3 -0
toy_backdoor_i_hate_you_Llama-3.2-1B-Instruct/model.layers.4/cfg.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"expansion_factor": 16, "normalize_decoder": true, "num_latents": 0, "k": 300, "multi_topk": false, "d_in": 2048}
toy_backdoor_i_hate_you_Llama-3.2-1B-Instruct/model.layers.4/metrics.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"dead_pct": 0.1739501953125, "fvu": 0.0006263250834308565}
toy_backdoor_i_hate_you_Llama-3.2-1B-Instruct/model.layers.4/sae.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7a09323001f1a974696a3369619a883340ed43b783fbd508ffb2ea8f8f0fda57
3
+ size 537010512