Hugging Face's logo

canrager
/
saebench_gemma-2-2b_width-2pow16_date-1109

Model card Files Community
canrager commited on
Commit
9f0d5d9
·
verified ·
1 Parent(s): d737e4e

Delete gemma-2-2b_topk_width-2pow16_date-1109/resid_post_layer_19_checkpoints

Browse files
This view is limited to 50 files because it contains too many changes.   See raw diff
Files changed (50) hide show
  1. gemma-2-2b_topk_width-2pow16_date-1109/resid_post_layer_19_checkpoints/trainer_0_step_0/ae.pt +0 -3
  2. gemma-2-2b_topk_width-2pow16_date-1109/resid_post_layer_19_checkpoints/trainer_0_step_0/config.json +0 -26
  3. gemma-2-2b_topk_width-2pow16_date-1109/resid_post_layer_19_checkpoints/trainer_0_step_195/ae.pt +0 -3
  4. gemma-2-2b_topk_width-2pow16_date-1109/resid_post_layer_19_checkpoints/trainer_0_step_195/config.json +0 -26
  5. gemma-2-2b_topk_width-2pow16_date-1109/resid_post_layer_19_checkpoints/trainer_0_step_1953/config.json +0 -26
  6. gemma-2-2b_topk_width-2pow16_date-1109/resid_post_layer_19_checkpoints/trainer_0_step_19531/config.json +0 -26
  7. gemma-2-2b_topk_width-2pow16_date-1109/resid_post_layer_19_checkpoints/trainer_0_step_617/config.json +0 -26
  8. gemma-2-2b_topk_width-2pow16_date-1109/resid_post_layer_19_checkpoints/trainer_0_step_6176/config.json +0 -26
  9. gemma-2-2b_topk_width-2pow16_date-1109/resid_post_layer_19_checkpoints/trainer_0_step_61763/config.json +0 -26
  10. gemma-2-2b_topk_width-2pow16_date-1109/resid_post_layer_19_checkpoints/trainer_1_step_0/ae.pt +0 -3
  11. gemma-2-2b_topk_width-2pow16_date-1109/resid_post_layer_19_checkpoints/trainer_1_step_0/config.json +0 -26
  12. gemma-2-2b_topk_width-2pow16_date-1109/resid_post_layer_19_checkpoints/trainer_1_step_195/ae.pt +0 -3
  13. gemma-2-2b_topk_width-2pow16_date-1109/resid_post_layer_19_checkpoints/trainer_1_step_195/config.json +0 -26
  14. gemma-2-2b_topk_width-2pow16_date-1109/resid_post_layer_19_checkpoints/trainer_1_step_1953/ae.pt +0 -3
  15. gemma-2-2b_topk_width-2pow16_date-1109/resid_post_layer_19_checkpoints/trainer_1_step_1953/config.json +0 -26
  16. gemma-2-2b_topk_width-2pow16_date-1109/resid_post_layer_19_checkpoints/trainer_1_step_19531/ae.pt +0 -3
  17. gemma-2-2b_topk_width-2pow16_date-1109/resid_post_layer_19_checkpoints/trainer_1_step_19531/config.json +0 -26
  18. gemma-2-2b_topk_width-2pow16_date-1109/resid_post_layer_19_checkpoints/trainer_1_step_617/ae.pt +0 -3
  19. gemma-2-2b_topk_width-2pow16_date-1109/resid_post_layer_19_checkpoints/trainer_1_step_617/config.json +0 -26
  20. gemma-2-2b_topk_width-2pow16_date-1109/resid_post_layer_19_checkpoints/trainer_1_step_6176/ae.pt +0 -3
  21. gemma-2-2b_topk_width-2pow16_date-1109/resid_post_layer_19_checkpoints/trainer_1_step_6176/config.json +0 -26
  22. gemma-2-2b_topk_width-2pow16_date-1109/resid_post_layer_19_checkpoints/trainer_1_step_61763/ae.pt +0 -3
  23. gemma-2-2b_topk_width-2pow16_date-1109/resid_post_layer_19_checkpoints/trainer_1_step_61763/config.json +0 -26
  24. gemma-2-2b_topk_width-2pow16_date-1109/resid_post_layer_19_checkpoints/trainer_2_step_0/ae.pt +0 -3
  25. gemma-2-2b_topk_width-2pow16_date-1109/resid_post_layer_19_checkpoints/trainer_2_step_0/config.json +0 -26
  26. gemma-2-2b_topk_width-2pow16_date-1109/resid_post_layer_19_checkpoints/trainer_2_step_195/ae.pt +0 -3
  27. gemma-2-2b_topk_width-2pow16_date-1109/resid_post_layer_19_checkpoints/trainer_2_step_195/config.json +0 -26
  28. gemma-2-2b_topk_width-2pow16_date-1109/resid_post_layer_19_checkpoints/trainer_2_step_1953/ae.pt +0 -3
  29. gemma-2-2b_topk_width-2pow16_date-1109/resid_post_layer_19_checkpoints/trainer_2_step_1953/config.json +0 -26
  30. gemma-2-2b_topk_width-2pow16_date-1109/resid_post_layer_19_checkpoints/trainer_2_step_19531/ae.pt +0 -3
  31. gemma-2-2b_topk_width-2pow16_date-1109/resid_post_layer_19_checkpoints/trainer_2_step_19531/config.json +0 -26
  32. gemma-2-2b_topk_width-2pow16_date-1109/resid_post_layer_19_checkpoints/trainer_2_step_617/ae.pt +0 -3
  33. gemma-2-2b_topk_width-2pow16_date-1109/resid_post_layer_19_checkpoints/trainer_2_step_617/config.json +0 -26
  34. gemma-2-2b_topk_width-2pow16_date-1109/resid_post_layer_19_checkpoints/trainer_2_step_6176/ae.pt +0 -3
  35. gemma-2-2b_topk_width-2pow16_date-1109/resid_post_layer_19_checkpoints/trainer_2_step_6176/config.json +0 -26
  36. gemma-2-2b_topk_width-2pow16_date-1109/resid_post_layer_19_checkpoints/trainer_2_step_61763/ae.pt +0 -3
  37. gemma-2-2b_topk_width-2pow16_date-1109/resid_post_layer_19_checkpoints/trainer_2_step_61763/config.json +0 -26
  38. gemma-2-2b_topk_width-2pow16_date-1109/resid_post_layer_19_checkpoints/trainer_3_step_0/ae.pt +0 -3
  39. gemma-2-2b_topk_width-2pow16_date-1109/resid_post_layer_19_checkpoints/trainer_3_step_0/config.json +0 -26
  40. gemma-2-2b_topk_width-2pow16_date-1109/resid_post_layer_19_checkpoints/trainer_3_step_195/ae.pt +0 -3
  41. gemma-2-2b_topk_width-2pow16_date-1109/resid_post_layer_19_checkpoints/trainer_3_step_195/config.json +0 -26
  42. gemma-2-2b_topk_width-2pow16_date-1109/resid_post_layer_19_checkpoints/trainer_3_step_1953/ae.pt +0 -3
  43. gemma-2-2b_topk_width-2pow16_date-1109/resid_post_layer_19_checkpoints/trainer_3_step_1953/config.json +0 -26
  44. gemma-2-2b_topk_width-2pow16_date-1109/resid_post_layer_19_checkpoints/trainer_3_step_19531/ae.pt +0 -3
  45. gemma-2-2b_topk_width-2pow16_date-1109/resid_post_layer_19_checkpoints/trainer_3_step_19531/config.json +0 -26
  46. gemma-2-2b_topk_width-2pow16_date-1109/resid_post_layer_19_checkpoints/trainer_3_step_617/ae.pt +0 -3
  47. gemma-2-2b_topk_width-2pow16_date-1109/resid_post_layer_19_checkpoints/trainer_3_step_617/config.json +0 -26
  48. gemma-2-2b_topk_width-2pow16_date-1109/resid_post_layer_19_checkpoints/trainer_3_step_6176/ae.pt +0 -3
  49. gemma-2-2b_topk_width-2pow16_date-1109/resid_post_layer_19_checkpoints/trainer_3_step_6176/config.json +0 -26
  50. gemma-2-2b_topk_width-2pow16_date-1109/resid_post_layer_19_checkpoints/trainer_3_step_61763/ae.pt +0 -3
gemma-2-2b_topk_width-2pow16_date-1109/resid_post_layer_19_checkpoints/trainer_0_step_0/ae.pt DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:9882317ab593dbcdc4c1289214117307d075cff2fe9c60af6b59cb1009c8b014
3
- size 1189354296
 
 
 
 
gemma-2-2b_topk_width-2pow16_date-1109/resid_post_layer_19_checkpoints/trainer_0_step_0/config.json DELETED
@@ -1,26 +0,0 @@
1
- {
2
- "trainer": {
3
- "trainer_class": "TrainerTopK",
4
- "dict_class": "AutoEncoderTopK",
5
- "lr": 0.00010079052613579392,
6
- "steps": "0",
7
- "seed": 0,
8
- "activation_dim": 2304,
9
- "dict_size": 64512,
10
- "k": 20,
11
- "device": "cuda:0",
12
- "layer": 19,
13
- "lm_name": "google/gemma-2-2b",
14
- "wandb_name": "TopKTrainer-google/gemma-2-2b-resid_post_layer_19",
15
- "submodule_name": "resid_post_layer_19"
16
- },
17
- "buffer": {
18
- "d_submodule": 2304,
19
- "io": "out",
20
- "n_ctxs": 2000,
21
- "ctx_len": 128,
22
- "refresh_batch_size": 24,
23
- "out_batch_size": 2048,
24
- "device": "cuda:0"
25
- }
26
- }
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
gemma-2-2b_topk_width-2pow16_date-1109/resid_post_layer_19_checkpoints/trainer_0_step_195/ae.pt DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:ae513ee96d00f991d178c3175705385db954be2ace3c14c55d8fe5afe53e1240
3
- size 1189354312
 
 
 
 
gemma-2-2b_topk_width-2pow16_date-1109/resid_post_layer_19_checkpoints/trainer_0_step_195/config.json DELETED
@@ -1,26 +0,0 @@
1
- {
2
- "trainer": {
3
- "trainer_class": "TrainerTopK",
4
- "dict_class": "AutoEncoderTopK",
5
- "lr": 0.00010079052613579392,
6
- "steps": "195",
7
- "seed": 0,
8
- "activation_dim": 2304,
9
- "dict_size": 64512,
10
- "k": 20,
11
- "device": "cuda:0",
12
- "layer": 19,
13
- "lm_name": "google/gemma-2-2b",
14
- "wandb_name": "TopKTrainer-google/gemma-2-2b-resid_post_layer_19",
15
- "submodule_name": "resid_post_layer_19"
16
- },
17
- "buffer": {
18
- "d_submodule": 2304,
19
- "io": "out",
20
- "n_ctxs": 2000,
21
- "ctx_len": 128,
22
- "refresh_batch_size": 24,
23
- "out_batch_size": 2048,
24
- "device": "cuda:0"
25
- }
26
- }
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
gemma-2-2b_topk_width-2pow16_date-1109/resid_post_layer_19_checkpoints/trainer_0_step_1953/config.json DELETED
@@ -1,26 +0,0 @@
1
- {
2
- "trainer": {
3
- "trainer_class": "TrainerTopK",
4
- "dict_class": "AutoEncoderTopK",
5
- "lr": 0.00010079052613579392,
6
- "steps": "1953",
7
- "seed": 0,
8
- "activation_dim": 2304,
9
- "dict_size": 64512,
10
- "k": 20,
11
- "device": "cuda:0",
12
- "layer": 19,
13
- "lm_name": "google/gemma-2-2b",
14
- "wandb_name": "TopKTrainer-google/gemma-2-2b-resid_post_layer_19",
15
- "submodule_name": "resid_post_layer_19"
16
- },
17
- "buffer": {
18
- "d_submodule": 2304,
19
- "io": "out",
20
- "n_ctxs": 2000,
21
- "ctx_len": 128,
22
- "refresh_batch_size": 24,
23
- "out_batch_size": 2048,
24
- "device": "cuda:0"
25
- }
26
- }
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
gemma-2-2b_topk_width-2pow16_date-1109/resid_post_layer_19_checkpoints/trainer_0_step_19531/config.json DELETED
@@ -1,26 +0,0 @@
1
- {
2
- "trainer": {
3
- "trainer_class": "TrainerTopK",
4
- "dict_class": "AutoEncoderTopK",
5
- "lr": 0.00010079052613579392,
6
- "steps": "19531",
7
- "seed": 0,
8
- "activation_dim": 2304,
9
- "dict_size": 64512,
10
- "k": 20,
11
- "device": "cuda:0",
12
- "layer": 19,
13
- "lm_name": "google/gemma-2-2b",
14
- "wandb_name": "TopKTrainer-google/gemma-2-2b-resid_post_layer_19",
15
- "submodule_name": "resid_post_layer_19"
16
- },
17
- "buffer": {
18
- "d_submodule": 2304,
19
- "io": "out",
20
- "n_ctxs": 2000,
21
- "ctx_len": 128,
22
- "refresh_batch_size": 24,
23
- "out_batch_size": 2048,
24
- "device": "cuda:0"
25
- }
26
- }
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
gemma-2-2b_topk_width-2pow16_date-1109/resid_post_layer_19_checkpoints/trainer_0_step_617/config.json DELETED
@@ -1,26 +0,0 @@
1
- {
2
- "trainer": {
3
- "trainer_class": "TrainerTopK",
4
- "dict_class": "AutoEncoderTopK",
5
- "lr": 0.00010079052613579392,
6
- "steps": "617",
7
- "seed": 0,
8
- "activation_dim": 2304,
9
- "dict_size": 64512,
10
- "k": 20,
11
- "device": "cuda:0",
12
- "layer": 19,
13
- "lm_name": "google/gemma-2-2b",
14
- "wandb_name": "TopKTrainer-google/gemma-2-2b-resid_post_layer_19",
15
- "submodule_name": "resid_post_layer_19"
16
- },
17
- "buffer": {
18
- "d_submodule": 2304,
19
- "io": "out",
20
- "n_ctxs": 2000,
21
- "ctx_len": 128,
22
- "refresh_batch_size": 24,
23
- "out_batch_size": 2048,
24
- "device": "cuda:0"
25
- }
26
- }
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
gemma-2-2b_topk_width-2pow16_date-1109/resid_post_layer_19_checkpoints/trainer_0_step_6176/config.json DELETED
@@ -1,26 +0,0 @@
1
- {
2
- "trainer": {
3
- "trainer_class": "TrainerTopK",
4
- "dict_class": "AutoEncoderTopK",
5
- "lr": 0.00010079052613579392,
6
- "steps": "6176",
7
- "seed": 0,
8
- "activation_dim": 2304,
9
- "dict_size": 64512,
10
- "k": 20,
11
- "device": "cuda:0",
12
- "layer": 19,
13
- "lm_name": "google/gemma-2-2b",
14
- "wandb_name": "TopKTrainer-google/gemma-2-2b-resid_post_layer_19",
15
- "submodule_name": "resid_post_layer_19"
16
- },
17
- "buffer": {
18
- "d_submodule": 2304,
19
- "io": "out",
20
- "n_ctxs": 2000,
21
- "ctx_len": 128,
22
- "refresh_batch_size": 24,
23
- "out_batch_size": 2048,
24
- "device": "cuda:0"
25
- }
26
- }
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
gemma-2-2b_topk_width-2pow16_date-1109/resid_post_layer_19_checkpoints/trainer_0_step_61763/config.json DELETED
@@ -1,26 +0,0 @@
1
- {
2
- "trainer": {
3
- "trainer_class": "TrainerTopK",
4
- "dict_class": "AutoEncoderTopK",
5
- "lr": 0.00010079052613579392,
6
- "steps": "61763",
7
- "seed": 0,
8
- "activation_dim": 2304,
9
- "dict_size": 64512,
10
- "k": 20,
11
- "device": "cuda:0",
12
- "layer": 19,
13
- "lm_name": "google/gemma-2-2b",
14
- "wandb_name": "TopKTrainer-google/gemma-2-2b-resid_post_layer_19",
15
- "submodule_name": "resid_post_layer_19"
16
- },
17
- "buffer": {
18
- "d_submodule": 2304,
19
- "io": "out",
20
- "n_ctxs": 2000,
21
- "ctx_len": 128,
22
- "refresh_batch_size": 24,
23
- "out_batch_size": 2048,
24
- "device": "cuda:0"
25
- }
26
- }
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
gemma-2-2b_topk_width-2pow16_date-1109/resid_post_layer_19_checkpoints/trainer_1_step_0/ae.pt DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:9882317ab593dbcdc4c1289214117307d075cff2fe9c60af6b59cb1009c8b014
3
- size 1189354296
 
 
 
 
gemma-2-2b_topk_width-2pow16_date-1109/resid_post_layer_19_checkpoints/trainer_1_step_0/config.json DELETED
@@ -1,26 +0,0 @@
1
- {
2
- "trainer": {
3
- "trainer_class": "TrainerTopK",
4
- "dict_class": "AutoEncoderTopK",
5
- "lr": 0.00010079052613579392,
6
- "steps": "0",
7
- "seed": 0,
8
- "activation_dim": 2304,
9
- "dict_size": 64512,
10
- "k": 40,
11
- "device": "cuda:0",
12
- "layer": 19,
13
- "lm_name": "google/gemma-2-2b",
14
- "wandb_name": "TopKTrainer-google/gemma-2-2b-resid_post_layer_19",
15
- "submodule_name": "resid_post_layer_19"
16
- },
17
- "buffer": {
18
- "d_submodule": 2304,
19
- "io": "out",
20
- "n_ctxs": 2000,
21
- "ctx_len": 128,
22
- "refresh_batch_size": 24,
23
- "out_batch_size": 2048,
24
- "device": "cuda:0"
25
- }
26
- }
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
gemma-2-2b_topk_width-2pow16_date-1109/resid_post_layer_19_checkpoints/trainer_1_step_195/ae.pt DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:9e0f5988fd15c6170506894e6309023c272e4c9aabe7cb807b434088799c6c67
3
- size 1189354312
 
 
 
 
gemma-2-2b_topk_width-2pow16_date-1109/resid_post_layer_19_checkpoints/trainer_1_step_195/config.json DELETED
@@ -1,26 +0,0 @@
1
- {
2
- "trainer": {
3
- "trainer_class": "TrainerTopK",
4
- "dict_class": "AutoEncoderTopK",
5
- "lr": 0.00010079052613579392,
6
- "steps": "195",
7
- "seed": 0,
8
- "activation_dim": 2304,
9
- "dict_size": 64512,
10
- "k": 40,
11
- "device": "cuda:0",
12
- "layer": 19,
13
- "lm_name": "google/gemma-2-2b",
14
- "wandb_name": "TopKTrainer-google/gemma-2-2b-resid_post_layer_19",
15
- "submodule_name": "resid_post_layer_19"
16
- },
17
- "buffer": {
18
- "d_submodule": 2304,
19
- "io": "out",
20
- "n_ctxs": 2000,
21
- "ctx_len": 128,
22
- "refresh_batch_size": 24,
23
- "out_batch_size": 2048,
24
- "device": "cuda:0"
25
- }
26
- }
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
gemma-2-2b_topk_width-2pow16_date-1109/resid_post_layer_19_checkpoints/trainer_1_step_1953/ae.pt DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:11ecb94de0079b249ace915f6fb558ef187a11b5fd463fb3a65ebe0dff5850ca
3
- size 1189354384
 
 
 
 
gemma-2-2b_topk_width-2pow16_date-1109/resid_post_layer_19_checkpoints/trainer_1_step_1953/config.json DELETED
@@ -1,26 +0,0 @@
1
- {
2
- "trainer": {
3
- "trainer_class": "TrainerTopK",
4
- "dict_class": "AutoEncoderTopK",
5
- "lr": 0.00010079052613579392,
6
- "steps": "1953",
7
- "seed": 0,
8
- "activation_dim": 2304,
9
- "dict_size": 64512,
10
- "k": 40,
11
- "device": "cuda:0",
12
- "layer": 19,
13
- "lm_name": "google/gemma-2-2b",
14
- "wandb_name": "TopKTrainer-google/gemma-2-2b-resid_post_layer_19",
15
- "submodule_name": "resid_post_layer_19"
16
- },
17
- "buffer": {
18
- "d_submodule": 2304,
19
- "io": "out",
20
- "n_ctxs": 2000,
21
- "ctx_len": 128,
22
- "refresh_batch_size": 24,
23
- "out_batch_size": 2048,
24
- "device": "cuda:0"
25
- }
26
- }
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
gemma-2-2b_topk_width-2pow16_date-1109/resid_post_layer_19_checkpoints/trainer_1_step_19531/ae.pt DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:d12aa6d1d0167c0cfa925610f1ea27c8fe56a5a382635f31c9d5e68211dec97a
3
- size 1189354584
 
 
 
 
gemma-2-2b_topk_width-2pow16_date-1109/resid_post_layer_19_checkpoints/trainer_1_step_19531/config.json DELETED
@@ -1,26 +0,0 @@
1
- {
2
- "trainer": {
3
- "trainer_class": "TrainerTopK",
4
- "dict_class": "AutoEncoderTopK",
5
- "lr": 0.00010079052613579392,
6
- "steps": "19531",
7
- "seed": 0,
8
- "activation_dim": 2304,
9
- "dict_size": 64512,
10
- "k": 40,
11
- "device": "cuda:0",
12
- "layer": 19,
13
- "lm_name": "google/gemma-2-2b",
14
- "wandb_name": "TopKTrainer-google/gemma-2-2b-resid_post_layer_19",
15
- "submodule_name": "resid_post_layer_19"
16
- },
17
- "buffer": {
18
- "d_submodule": 2304,
19
- "io": "out",
20
- "n_ctxs": 2000,
21
- "ctx_len": 128,
22
- "refresh_batch_size": 24,
23
- "out_batch_size": 2048,
24
- "device": "cuda:0"
25
- }
26
- }
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
gemma-2-2b_topk_width-2pow16_date-1109/resid_post_layer_19_checkpoints/trainer_1_step_617/ae.pt DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:ea8ec2b4131dec7f52d2772af0502e3658008b5f0cf35af2529d80256a12a350
3
- size 1189354312
 
 
 
 
gemma-2-2b_topk_width-2pow16_date-1109/resid_post_layer_19_checkpoints/trainer_1_step_617/config.json DELETED
@@ -1,26 +0,0 @@
1
- {
2
- "trainer": {
3
- "trainer_class": "TrainerTopK",
4
- "dict_class": "AutoEncoderTopK",
5
- "lr": 0.00010079052613579392,
6
- "steps": "617",
7
- "seed": 0,
8
- "activation_dim": 2304,
9
- "dict_size": 64512,
10
- "k": 40,
11
- "device": "cuda:0",
12
- "layer": 19,
13
- "lm_name": "google/gemma-2-2b",
14
- "wandb_name": "TopKTrainer-google/gemma-2-2b-resid_post_layer_19",
15
- "submodule_name": "resid_post_layer_19"
16
- },
17
- "buffer": {
18
- "d_submodule": 2304,
19
- "io": "out",
20
- "n_ctxs": 2000,
21
- "ctx_len": 128,
22
- "refresh_batch_size": 24,
23
- "out_batch_size": 2048,
24
- "device": "cuda:0"
25
- }
26
- }
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
gemma-2-2b_topk_width-2pow16_date-1109/resid_post_layer_19_checkpoints/trainer_1_step_6176/ae.pt DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:5169cb2decddc91af20ec1d2aa1681bb7c9d75f74ce4052c72d7d6c6b2305fad
3
- size 1189354384
 
 
 
 
gemma-2-2b_topk_width-2pow16_date-1109/resid_post_layer_19_checkpoints/trainer_1_step_6176/config.json DELETED
@@ -1,26 +0,0 @@
1
- {
2
- "trainer": {
3
- "trainer_class": "TrainerTopK",
4
- "dict_class": "AutoEncoderTopK",
5
- "lr": 0.00010079052613579392,
6
- "steps": "6176",
7
- "seed": 0,
8
- "activation_dim": 2304,
9
- "dict_size": 64512,
10
- "k": 40,
11
- "device": "cuda:0",
12
- "layer": 19,
13
- "lm_name": "google/gemma-2-2b",
14
- "wandb_name": "TopKTrainer-google/gemma-2-2b-resid_post_layer_19",
15
- "submodule_name": "resid_post_layer_19"
16
- },
17
- "buffer": {
18
- "d_submodule": 2304,
19
- "io": "out",
20
- "n_ctxs": 2000,
21
- "ctx_len": 128,
22
- "refresh_batch_size": 24,
23
- "out_batch_size": 2048,
24
- "device": "cuda:0"
25
- }
26
- }
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
gemma-2-2b_topk_width-2pow16_date-1109/resid_post_layer_19_checkpoints/trainer_1_step_61763/ae.pt DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:a1d7aeef9a3f42b21329ff32e0f6b3b3092d71e36d7c455c3c402186b1ddf1d9
3
- size 1189354584
 
 
 
 
gemma-2-2b_topk_width-2pow16_date-1109/resid_post_layer_19_checkpoints/trainer_1_step_61763/config.json DELETED
@@ -1,26 +0,0 @@
1
- {
2
- "trainer": {
3
- "trainer_class": "TrainerTopK",
4
- "dict_class": "AutoEncoderTopK",
5
- "lr": 0.00010079052613579392,
6
- "steps": "61763",
7
- "seed": 0,
8
- "activation_dim": 2304,
9
- "dict_size": 64512,
10
- "k": 40,
11
- "device": "cuda:0",
12
- "layer": 19,
13
- "lm_name": "google/gemma-2-2b",
14
- "wandb_name": "TopKTrainer-google/gemma-2-2b-resid_post_layer_19",
15
- "submodule_name": "resid_post_layer_19"
16
- },
17
- "buffer": {
18
- "d_submodule": 2304,
19
- "io": "out",
20
- "n_ctxs": 2000,
21
- "ctx_len": 128,
22
- "refresh_batch_size": 24,
23
- "out_batch_size": 2048,
24
- "device": "cuda:0"
25
- }
26
- }
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
gemma-2-2b_topk_width-2pow16_date-1109/resid_post_layer_19_checkpoints/trainer_2_step_0/ae.pt DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:9882317ab593dbcdc4c1289214117307d075cff2fe9c60af6b59cb1009c8b014
3
- size 1189354296
 
 
 
 
gemma-2-2b_topk_width-2pow16_date-1109/resid_post_layer_19_checkpoints/trainer_2_step_0/config.json DELETED
@@ -1,26 +0,0 @@
1
- {
2
- "trainer": {
3
- "trainer_class": "TrainerTopK",
4
- "dict_class": "AutoEncoderTopK",
5
- "lr": 0.00010079052613579392,
6
- "steps": "0",
7
- "seed": 0,
8
- "activation_dim": 2304,
9
- "dict_size": 64512,
10
- "k": 80,
11
- "device": "cuda:0",
12
- "layer": 19,
13
- "lm_name": "google/gemma-2-2b",
14
- "wandb_name": "TopKTrainer-google/gemma-2-2b-resid_post_layer_19",
15
- "submodule_name": "resid_post_layer_19"
16
- },
17
- "buffer": {
18
- "d_submodule": 2304,
19
- "io": "out",
20
- "n_ctxs": 2000,
21
- "ctx_len": 128,
22
- "refresh_batch_size": 24,
23
- "out_batch_size": 2048,
24
- "device": "cuda:0"
25
- }
26
- }
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
gemma-2-2b_topk_width-2pow16_date-1109/resid_post_layer_19_checkpoints/trainer_2_step_195/ae.pt DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:0b5f543a8b790108264e9f11a9bc5f39e4246abca94d7ff43339d7245acc2f0e
3
- size 1189354312
 
 
 
 
gemma-2-2b_topk_width-2pow16_date-1109/resid_post_layer_19_checkpoints/trainer_2_step_195/config.json DELETED
@@ -1,26 +0,0 @@
1
- {
2
- "trainer": {
3
- "trainer_class": "TrainerTopK",
4
- "dict_class": "AutoEncoderTopK",
5
- "lr": 0.00010079052613579392,
6
- "steps": "195",
7
- "seed": 0,
8
- "activation_dim": 2304,
9
- "dict_size": 64512,
10
- "k": 80,
11
- "device": "cuda:0",
12
- "layer": 19,
13
- "lm_name": "google/gemma-2-2b",
14
- "wandb_name": "TopKTrainer-google/gemma-2-2b-resid_post_layer_19",
15
- "submodule_name": "resid_post_layer_19"
16
- },
17
- "buffer": {
18
- "d_submodule": 2304,
19
- "io": "out",
20
- "n_ctxs": 2000,
21
- "ctx_len": 128,
22
- "refresh_batch_size": 24,
23
- "out_batch_size": 2048,
24
- "device": "cuda:0"
25
- }
26
- }
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
gemma-2-2b_topk_width-2pow16_date-1109/resid_post_layer_19_checkpoints/trainer_2_step_1953/ae.pt DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:1c98dc746b35b2de998cd9f3c2c1bfd483ef2bdc789f60bcb1b041acdde97917
3
- size 1189354384
 
 
 
 
gemma-2-2b_topk_width-2pow16_date-1109/resid_post_layer_19_checkpoints/trainer_2_step_1953/config.json DELETED
@@ -1,26 +0,0 @@
1
- {
2
- "trainer": {
3
- "trainer_class": "TrainerTopK",
4
- "dict_class": "AutoEncoderTopK",
5
- "lr": 0.00010079052613579392,
6
- "steps": "1953",
7
- "seed": 0,
8
- "activation_dim": 2304,
9
- "dict_size": 64512,
10
- "k": 80,
11
- "device": "cuda:0",
12
- "layer": 19,
13
- "lm_name": "google/gemma-2-2b",
14
- "wandb_name": "TopKTrainer-google/gemma-2-2b-resid_post_layer_19",
15
- "submodule_name": "resid_post_layer_19"
16
- },
17
- "buffer": {
18
- "d_submodule": 2304,
19
- "io": "out",
20
- "n_ctxs": 2000,
21
- "ctx_len": 128,
22
- "refresh_batch_size": 24,
23
- "out_batch_size": 2048,
24
- "device": "cuda:0"
25
- }
26
- }
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
gemma-2-2b_topk_width-2pow16_date-1109/resid_post_layer_19_checkpoints/trainer_2_step_19531/ae.pt DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:6a4ed0c6f1c643f7857f1ed5863f02881d061cb2664272488cbbe6b1a995bd78
3
- size 1189354584
 
 
 
 
gemma-2-2b_topk_width-2pow16_date-1109/resid_post_layer_19_checkpoints/trainer_2_step_19531/config.json DELETED
@@ -1,26 +0,0 @@
1
- {
2
- "trainer": {
3
- "trainer_class": "TrainerTopK",
4
- "dict_class": "AutoEncoderTopK",
5
- "lr": 0.00010079052613579392,
6
- "steps": "19531",
7
- "seed": 0,
8
- "activation_dim": 2304,
9
- "dict_size": 64512,
10
- "k": 80,
11
- "device": "cuda:0",
12
- "layer": 19,
13
- "lm_name": "google/gemma-2-2b",
14
- "wandb_name": "TopKTrainer-google/gemma-2-2b-resid_post_layer_19",
15
- "submodule_name": "resid_post_layer_19"
16
- },
17
- "buffer": {
18
- "d_submodule": 2304,
19
- "io": "out",
20
- "n_ctxs": 2000,
21
- "ctx_len": 128,
22
- "refresh_batch_size": 24,
23
- "out_batch_size": 2048,
24
- "device": "cuda:0"
25
- }
26
- }
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
gemma-2-2b_topk_width-2pow16_date-1109/resid_post_layer_19_checkpoints/trainer_2_step_617/ae.pt DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:8f5a58bec7b367bed4f09c06fc8f6bcc96f9f78d8fea6812979aaab284ade153
3
- size 1189354312
 
 
 
 
gemma-2-2b_topk_width-2pow16_date-1109/resid_post_layer_19_checkpoints/trainer_2_step_617/config.json DELETED
@@ -1,26 +0,0 @@
1
- {
2
- "trainer": {
3
- "trainer_class": "TrainerTopK",
4
- "dict_class": "AutoEncoderTopK",
5
- "lr": 0.00010079052613579392,
6
- "steps": "617",
7
- "seed": 0,
8
- "activation_dim": 2304,
9
- "dict_size": 64512,
10
- "k": 80,
11
- "device": "cuda:0",
12
- "layer": 19,
13
- "lm_name": "google/gemma-2-2b",
14
- "wandb_name": "TopKTrainer-google/gemma-2-2b-resid_post_layer_19",
15
- "submodule_name": "resid_post_layer_19"
16
- },
17
- "buffer": {
18
- "d_submodule": 2304,
19
- "io": "out",
20
- "n_ctxs": 2000,
21
- "ctx_len": 128,
22
- "refresh_batch_size": 24,
23
- "out_batch_size": 2048,
24
- "device": "cuda:0"
25
- }
26
- }
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
gemma-2-2b_topk_width-2pow16_date-1109/resid_post_layer_19_checkpoints/trainer_2_step_6176/ae.pt DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:04d8b0de11a5d09081caf84c4b823e891a6c2a6977acf5e1c9a25f845dc05b7a
3
- size 1189354384
 
 
 
 
gemma-2-2b_topk_width-2pow16_date-1109/resid_post_layer_19_checkpoints/trainer_2_step_6176/config.json DELETED
@@ -1,26 +0,0 @@
1
- {
2
- "trainer": {
3
- "trainer_class": "TrainerTopK",
4
- "dict_class": "AutoEncoderTopK",
5
- "lr": 0.00010079052613579392,
6
- "steps": "6176",
7
- "seed": 0,
8
- "activation_dim": 2304,
9
- "dict_size": 64512,
10
- "k": 80,
11
- "device": "cuda:0",
12
- "layer": 19,
13
- "lm_name": "google/gemma-2-2b",
14
- "wandb_name": "TopKTrainer-google/gemma-2-2b-resid_post_layer_19",
15
- "submodule_name": "resid_post_layer_19"
16
- },
17
- "buffer": {
18
- "d_submodule": 2304,
19
- "io": "out",
20
- "n_ctxs": 2000,
21
- "ctx_len": 128,
22
- "refresh_batch_size": 24,
23
- "out_batch_size": 2048,
24
- "device": "cuda:0"
25
- }
26
- }
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
gemma-2-2b_topk_width-2pow16_date-1109/resid_post_layer_19_checkpoints/trainer_2_step_61763/ae.pt DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:c65b8a48410eec893b0937ebb297ab23dabea2eea039ab0b109adacd02c42a06
3
- size 1189354584
 
 
 
 
gemma-2-2b_topk_width-2pow16_date-1109/resid_post_layer_19_checkpoints/trainer_2_step_61763/config.json DELETED
@@ -1,26 +0,0 @@
1
- {
2
- "trainer": {
3
- "trainer_class": "TrainerTopK",
4
- "dict_class": "AutoEncoderTopK",
5
- "lr": 0.00010079052613579392,
6
- "steps": "61763",
7
- "seed": 0,
8
- "activation_dim": 2304,
9
- "dict_size": 64512,
10
- "k": 80,
11
- "device": "cuda:0",
12
- "layer": 19,
13
- "lm_name": "google/gemma-2-2b",
14
- "wandb_name": "TopKTrainer-google/gemma-2-2b-resid_post_layer_19",
15
- "submodule_name": "resid_post_layer_19"
16
- },
17
- "buffer": {
18
- "d_submodule": 2304,
19
- "io": "out",
20
- "n_ctxs": 2000,
21
- "ctx_len": 128,
22
- "refresh_batch_size": 24,
23
- "out_batch_size": 2048,
24
- "device": "cuda:0"
25
- }
26
- }
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
gemma-2-2b_topk_width-2pow16_date-1109/resid_post_layer_19_checkpoints/trainer_3_step_0/ae.pt DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:9882317ab593dbcdc4c1289214117307d075cff2fe9c60af6b59cb1009c8b014
3
- size 1189354296
 
 
 
 
gemma-2-2b_topk_width-2pow16_date-1109/resid_post_layer_19_checkpoints/trainer_3_step_0/config.json DELETED
@@ -1,26 +0,0 @@
1
- {
2
- "trainer": {
3
- "trainer_class": "TrainerTopK",
4
- "dict_class": "AutoEncoderTopK",
5
- "lr": 0.00010079052613579392,
6
- "steps": "0",
7
- "seed": 0,
8
- "activation_dim": 2304,
9
- "dict_size": 64512,
10
- "k": 160,
11
- "device": "cuda:0",
12
- "layer": 19,
13
- "lm_name": "google/gemma-2-2b",
14
- "wandb_name": "TopKTrainer-google/gemma-2-2b-resid_post_layer_19",
15
- "submodule_name": "resid_post_layer_19"
16
- },
17
- "buffer": {
18
- "d_submodule": 2304,
19
- "io": "out",
20
- "n_ctxs": 2000,
21
- "ctx_len": 128,
22
- "refresh_batch_size": 24,
23
- "out_batch_size": 2048,
24
- "device": "cuda:0"
25
- }
26
- }
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
gemma-2-2b_topk_width-2pow16_date-1109/resid_post_layer_19_checkpoints/trainer_3_step_195/ae.pt DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:71fb799dd4d8d998590171679863d0d7177a214a78563cb687ef62821bf4a7a8
3
- size 1189354312
 
 
 
 
gemma-2-2b_topk_width-2pow16_date-1109/resid_post_layer_19_checkpoints/trainer_3_step_195/config.json DELETED
@@ -1,26 +0,0 @@
1
- {
2
- "trainer": {
3
- "trainer_class": "TrainerTopK",
4
- "dict_class": "AutoEncoderTopK",
5
- "lr": 0.00010079052613579392,
6
- "steps": "195",
7
- "seed": 0,
8
- "activation_dim": 2304,
9
- "dict_size": 64512,
10
- "k": 160,
11
- "device": "cuda:0",
12
- "layer": 19,
13
- "lm_name": "google/gemma-2-2b",
14
- "wandb_name": "TopKTrainer-google/gemma-2-2b-resid_post_layer_19",
15
- "submodule_name": "resid_post_layer_19"
16
- },
17
- "buffer": {
18
- "d_submodule": 2304,
19
- "io": "out",
20
- "n_ctxs": 2000,
21
- "ctx_len": 128,
22
- "refresh_batch_size": 24,
23
- "out_batch_size": 2048,
24
- "device": "cuda:0"
25
- }
26
- }
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
gemma-2-2b_topk_width-2pow16_date-1109/resid_post_layer_19_checkpoints/trainer_3_step_1953/ae.pt DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:33b132d92e179c7887e837bb409d19ff97240b686b001e2091dac81e5601675b
3
- size 1189354384
 
 
 
 
gemma-2-2b_topk_width-2pow16_date-1109/resid_post_layer_19_checkpoints/trainer_3_step_1953/config.json DELETED
@@ -1,26 +0,0 @@
1
- {
2
- "trainer": {
3
- "trainer_class": "TrainerTopK",
4
- "dict_class": "AutoEncoderTopK",
5
- "lr": 0.00010079052613579392,
6
- "steps": "1953",
7
- "seed": 0,
8
- "activation_dim": 2304,
9
- "dict_size": 64512,
10
- "k": 160,
11
- "device": "cuda:0",
12
- "layer": 19,
13
- "lm_name": "google/gemma-2-2b",
14
- "wandb_name": "TopKTrainer-google/gemma-2-2b-resid_post_layer_19",
15
- "submodule_name": "resid_post_layer_19"
16
- },
17
- "buffer": {
18
- "d_submodule": 2304,
19
- "io": "out",
20
- "n_ctxs": 2000,
21
- "ctx_len": 128,
22
- "refresh_batch_size": 24,
23
- "out_batch_size": 2048,
24
- "device": "cuda:0"
25
- }
26
- }
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
gemma-2-2b_topk_width-2pow16_date-1109/resid_post_layer_19_checkpoints/trainer_3_step_19531/ae.pt DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:8cfbbcd14b061661c6351f3e166f8b13698160bde5b250b67b0b0b07795b0a09
3
- size 1189354584
 
 
 
 
gemma-2-2b_topk_width-2pow16_date-1109/resid_post_layer_19_checkpoints/trainer_3_step_19531/config.json DELETED
@@ -1,26 +0,0 @@
1
- {
2
- "trainer": {
3
- "trainer_class": "TrainerTopK",
4
- "dict_class": "AutoEncoderTopK",
5
- "lr": 0.00010079052613579392,
6
- "steps": "19531",
7
- "seed": 0,
8
- "activation_dim": 2304,
9
- "dict_size": 64512,
10
- "k": 160,
11
- "device": "cuda:0",
12
- "layer": 19,
13
- "lm_name": "google/gemma-2-2b",
14
- "wandb_name": "TopKTrainer-google/gemma-2-2b-resid_post_layer_19",
15
- "submodule_name": "resid_post_layer_19"
16
- },
17
- "buffer": {
18
- "d_submodule": 2304,
19
- "io": "out",
20
- "n_ctxs": 2000,
21
- "ctx_len": 128,
22
- "refresh_batch_size": 24,
23
- "out_batch_size": 2048,
24
- "device": "cuda:0"
25
- }
26
- }
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
gemma-2-2b_topk_width-2pow16_date-1109/resid_post_layer_19_checkpoints/trainer_3_step_617/ae.pt DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:6224021693529c16c3807f387250ee50a1cee07255b7bfb129072e45ddaf21db
3
- size 1189354312
 
 
 
 
gemma-2-2b_topk_width-2pow16_date-1109/resid_post_layer_19_checkpoints/trainer_3_step_617/config.json DELETED
@@ -1,26 +0,0 @@
1
- {
2
- "trainer": {
3
- "trainer_class": "TrainerTopK",
4
- "dict_class": "AutoEncoderTopK",
5
- "lr": 0.00010079052613579392,
6
- "steps": "617",
7
- "seed": 0,
8
- "activation_dim": 2304,
9
- "dict_size": 64512,
10
- "k": 160,
11
- "device": "cuda:0",
12
- "layer": 19,
13
- "lm_name": "google/gemma-2-2b",
14
- "wandb_name": "TopKTrainer-google/gemma-2-2b-resid_post_layer_19",
15
- "submodule_name": "resid_post_layer_19"
16
- },
17
- "buffer": {
18
- "d_submodule": 2304,
19
- "io": "out",
20
- "n_ctxs": 2000,
21
- "ctx_len": 128,
22
- "refresh_batch_size": 24,
23
- "out_batch_size": 2048,
24
- "device": "cuda:0"
25
- }
26
- }
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
gemma-2-2b_topk_width-2pow16_date-1109/resid_post_layer_19_checkpoints/trainer_3_step_6176/ae.pt DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:7e026a8e2e8a265e55f787540211dc8ff9ca74b534e1e9349ae9f7cba55f05a4
3
- size 1189354384
 
 
 
 
gemma-2-2b_topk_width-2pow16_date-1109/resid_post_layer_19_checkpoints/trainer_3_step_6176/config.json DELETED
@@ -1,26 +0,0 @@
1
- {
2
- "trainer": {
3
- "trainer_class": "TrainerTopK",
4
- "dict_class": "AutoEncoderTopK",
5
- "lr": 0.00010079052613579392,
6
- "steps": "6176",
7
- "seed": 0,
8
- "activation_dim": 2304,
9
- "dict_size": 64512,
10
- "k": 160,
11
- "device": "cuda:0",
12
- "layer": 19,
13
- "lm_name": "google/gemma-2-2b",
14
- "wandb_name": "TopKTrainer-google/gemma-2-2b-resid_post_layer_19",
15
- "submodule_name": "resid_post_layer_19"
16
- },
17
- "buffer": {
18
- "d_submodule": 2304,
19
- "io": "out",
20
- "n_ctxs": 2000,
21
- "ctx_len": 128,
22
- "refresh_batch_size": 24,
23
- "out_batch_size": 2048,
24
- "device": "cuda:0"
25
- }
26
- }
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
gemma-2-2b_topk_width-2pow16_date-1109/resid_post_layer_19_checkpoints/trainer_3_step_61763/ae.pt DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:f8d63b38228f299c9cf43ec402178f21dd6ec564cd91ff3bbd913f96e5b5486b
3
- size 1189354584