Abdullah
commited on
Uploading saes for ['transformer.h.0.attn', 'transformer.h.0.mlp', 'transformer.h.1.attn', 'transformer.h.1.mlp'] and sft_sql_interp_TinyStories-2Layers-33M_cs4_dataset_synonyms_experiment_1.1_syn=True/k=256
Browse files- saes_sft_sql_interp_TinyStories-2Layers-33M_cs4_dataset_synonyms_experiment_1.1_syn=True/k=256/model_config.json +1 -1
- saes_sft_sql_interp_TinyStories-2Layers-33M_cs4_dataset_synonyms_experiment_1.1_syn=True/k=256/optimizer.pt +1 -1
- saes_sft_sql_interp_TinyStories-2Layers-33M_cs4_dataset_synonyms_experiment_1.1_syn=True/k=256/state.pt +1 -1
- saes_sft_sql_interp_TinyStories-2Layers-33M_cs4_dataset_synonyms_experiment_1.1_syn=True/k=256/transformer.h.0.attn/metrics.json +1 -1
- saes_sft_sql_interp_TinyStories-2Layers-33M_cs4_dataset_synonyms_experiment_1.1_syn=True/k=256/transformer.h.0.attn/sae.safetensors +1 -1
- saes_sft_sql_interp_TinyStories-2Layers-33M_cs4_dataset_synonyms_experiment_1.1_syn=True/k=256/transformer.h.0.mlp/metrics.json +1 -1
- saes_sft_sql_interp_TinyStories-2Layers-33M_cs4_dataset_synonyms_experiment_1.1_syn=True/k=256/transformer.h.0.mlp/sae.safetensors +1 -1
- saes_sft_sql_interp_TinyStories-2Layers-33M_cs4_dataset_synonyms_experiment_1.1_syn=True/k=256/transformer.h.1.attn/metrics.json +1 -1
- saes_sft_sql_interp_TinyStories-2Layers-33M_cs4_dataset_synonyms_experiment_1.1_syn=True/k=256/transformer.h.1.attn/sae.safetensors +1 -1
- saes_sft_sql_interp_TinyStories-2Layers-33M_cs4_dataset_synonyms_experiment_1.1_syn=True/k=256/transformer.h.1.mlp/metrics.json +1 -1
- saes_sft_sql_interp_TinyStories-2Layers-33M_cs4_dataset_synonyms_experiment_1.1_syn=True/k=256/transformer.h.1.mlp/sae.safetensors +1 -1
saes_sft_sql_interp_TinyStories-2Layers-33M_cs4_dataset_synonyms_experiment_1.1_syn=True/k=256/model_config.json
CHANGED
|
@@ -1 +1 @@
|
|
| 1 |
-
{"model_name": "withmartian/sft_sql_interp_TinyStories-2Layers-33M_cs4_dataset_synonyms_experiment_1.1", "dataset_name": "withmartian/
|
|
|
|
| 1 |
+
{"model_name": "withmartian/sft_sql_interp_TinyStories-2Layers-33M_cs4_dataset_synonyms_experiment_1.1", "dataset_name": "withmartian/cs5_dataset_synonyms", "model_abbrev": "saes_sft_sql_interp_TinyStories-2Layers-33M_cs4_dataset_synonyms_experiment_1.1_syn=True"}
|
saes_sft_sql_interp_TinyStories-2Layers-33M_cs4_dataset_synonyms_experiment_1.1_syn=True/k=256/optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1007171903
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:70ba185d15c4266961c443120fd44653a290fd0daa91a72ae84e6189b611173e
|
| 3 |
size 1007171903
|
saes_sft_sql_interp_TinyStories-2Layers-33M_cs4_dataset_synonyms_experiment_1.1_syn=True/k=256/state.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 493769
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:53ad8da566ec46c2e7d8b4d63924ed1915bd5178e078270f57872c5896e24ff4
|
| 3 |
size 493769
|
saes_sft_sql_interp_TinyStories-2Layers-33M_cs4_dataset_synonyms_experiment_1.1_syn=True/k=256/transformer.h.0.attn/metrics.json
CHANGED
|
@@ -1 +1 @@
|
|
| 1 |
-
{"dead_pct": 0.
|
|
|
|
| 1 |
+
{"dead_pct": 0.004882812965661287, "fvu": 0.008325805887579918}
|
saes_sft_sql_interp_TinyStories-2Layers-33M_cs4_dataset_synonyms_experiment_1.1_syn=True/k=256/transformer.h.0.attn/sae.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 125894984
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:df97c5d6668086e484cdfa5d265334cfae49e09d402050fd290cca89f62d2177
|
| 3 |
size 125894984
|
saes_sft_sql_interp_TinyStories-2Layers-33M_cs4_dataset_synonyms_experiment_1.1_syn=True/k=256/transformer.h.0.mlp/metrics.json
CHANGED
|
@@ -1 +1 @@
|
|
| 1 |
-
{"dead_pct": 0.0, "fvu": 0.
|
|
|
|
| 1 |
+
{"dead_pct": 0.0, "fvu": 0.017450330778956413}
|
saes_sft_sql_interp_TinyStories-2Layers-33M_cs4_dataset_synonyms_experiment_1.1_syn=True/k=256/transformer.h.0.mlp/sae.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 125894984
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:cf130148630f3b36d1d9018438d6b19e3a234dfe8616713ad5b7eddebac6969c
|
| 3 |
size 125894984
|
saes_sft_sql_interp_TinyStories-2Layers-33M_cs4_dataset_synonyms_experiment_1.1_syn=True/k=256/transformer.h.1.attn/metrics.json
CHANGED
|
@@ -1 +1 @@
|
|
| 1 |
-
{"dead_pct": 0.
|
|
|
|
| 1 |
+
{"dead_pct": 0.07109375298023224, "fvu": 0.012197243049740791}
|
saes_sft_sql_interp_TinyStories-2Layers-33M_cs4_dataset_synonyms_experiment_1.1_syn=True/k=256/transformer.h.1.attn/sae.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 125894984
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:72efb21d8889eded524e33a71c8cb4f0335ffd02de87f8999d465f6475058e59
|
| 3 |
size 125894984
|
saes_sft_sql_interp_TinyStories-2Layers-33M_cs4_dataset_synonyms_experiment_1.1_syn=True/k=256/transformer.h.1.mlp/metrics.json
CHANGED
|
@@ -1 +1 @@
|
|
| 1 |
-
{"dead_pct": 0.0, "fvu": 0.
|
|
|
|
| 1 |
+
{"dead_pct": 0.0, "fvu": 0.021102389320731163}
|
saes_sft_sql_interp_TinyStories-2Layers-33M_cs4_dataset_synonyms_experiment_1.1_syn=True/k=256/transformer.h.1.mlp/sae.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 125894984
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:2fac135e9f7c8d17fc88afe8e3d8900f172fb7c6097b8e5e9255a40943e0b311
|
| 3 |
size 125894984
|