Hugging Face's logo

Datasets:
Exploration-Lab
/
dim-discovery-archive

ArXiv:
Dataset card Files
xet
 Community
Divyanshu-Bhatt commited on
Commit
1de942e
·
verified ·
1 Parent(s): 14cbfb5

Add files using upload-large-folder tool

Browse files
This view is limited to 50 files because it contains too many changes.   See raw diff
Files changed (50) hide show
  1. copa_train/few_shot_1/meta-llama_Llama-2-7b-chat-hf/last_token_relative_depth_100/accuracy_blocks.0.hook_mlp_out.json +1 -0
  2. copa_train/few_shot_1/meta-llama_Llama-2-7b-chat-hf/last_token_relative_depth_100/accuracy_blocks.11.hook_mlp_out.json +1 -0
  3. copa_train/few_shot_1/meta-llama_Llama-2-7b-chat-hf/last_token_relative_depth_100/accuracy_blocks.12.hook_mlp_out.json +1 -0
  4. copa_train/few_shot_1/meta-llama_Llama-2-7b-chat-hf/last_token_relative_depth_100/accuracy_blocks.12.hook_resid_post.json +1 -0
  5. copa_train/few_shot_1/meta-llama_Llama-2-7b-chat-hf/last_token_relative_depth_100/accuracy_blocks.13.hook_mlp_out.json +1 -0
  6. copa_train/few_shot_1/meta-llama_Llama-2-7b-chat-hf/last_token_relative_depth_100/accuracy_blocks.13.hook_resid_post.json +1 -0
  7. copa_train/few_shot_1/meta-llama_Llama-2-7b-chat-hf/last_token_relative_depth_100/accuracy_blocks.14.hook_mlp_out.json +1 -0
  8. copa_train/few_shot_1/meta-llama_Llama-2-7b-chat-hf/last_token_relative_depth_100/accuracy_blocks.15.hook_mlp_out.json +1 -0
  9. copa_train/few_shot_1/meta-llama_Llama-2-7b-chat-hf/last_token_relative_depth_100/accuracy_blocks.16.hook_mlp_out.json +1 -0
  10. copa_train/few_shot_1/meta-llama_Llama-2-7b-chat-hf/last_token_relative_depth_100/accuracy_blocks.16.hook_resid_post.json +1 -0
  11. copa_train/few_shot_1/meta-llama_Llama-2-7b-chat-hf/last_token_relative_depth_100/accuracy_blocks.17.hook_mlp_out.json +1 -0
  12. copa_train/few_shot_1/meta-llama_Llama-2-7b-chat-hf/last_token_relative_depth_100/accuracy_blocks.17.hook_resid_post.json +1 -0
  13. copa_train/few_shot_1/meta-llama_Llama-2-7b-chat-hf/last_token_relative_depth_100/accuracy_blocks.18.hook_resid_post.json +1 -0
  14. copa_train/few_shot_1/meta-llama_Llama-2-7b-chat-hf/last_token_relative_depth_100/accuracy_blocks.19.hook_resid_post.json +1 -0
  15. copa_train/few_shot_1/meta-llama_Llama-2-7b-chat-hf/last_token_relative_depth_100/accuracy_blocks.2.hook_mlp_out.json +1 -0
  16. copa_train/few_shot_1/meta-llama_Llama-2-7b-chat-hf/last_token_relative_depth_100/accuracy_blocks.2.hook_resid_post.json +1 -0
  17. copa_train/few_shot_1/meta-llama_Llama-2-7b-chat-hf/last_token_relative_depth_100/accuracy_blocks.20.hook_mlp_out.json +1 -0
  18. copa_train/few_shot_1/meta-llama_Llama-2-7b-chat-hf/last_token_relative_depth_100/accuracy_blocks.21.hook_mlp_out.json +1 -0
  19. copa_train/few_shot_1/meta-llama_Llama-2-7b-chat-hf/last_token_relative_depth_100/accuracy_blocks.21.hook_resid_post.json +1 -0
  20. copa_train/few_shot_1/meta-llama_Llama-2-7b-chat-hf/last_token_relative_depth_100/accuracy_blocks.22.hook_mlp_out.json +1 -0
  21. copa_train/few_shot_1/meta-llama_Llama-2-7b-chat-hf/last_token_relative_depth_100/accuracy_blocks.23.hook_mlp_out.json +1 -0
  22. copa_train/few_shot_1/meta-llama_Llama-2-7b-chat-hf/last_token_relative_depth_100/accuracy_blocks.24.hook_mlp_out.json +1 -0
  23. copa_train/few_shot_1/meta-llama_Llama-2-7b-chat-hf/last_token_relative_depth_100/accuracy_blocks.24.hook_resid_post.json +1 -0
  24. copa_train/few_shot_1/meta-llama_Llama-2-7b-chat-hf/last_token_relative_depth_100/accuracy_blocks.25.hook_mlp_out.json +1 -0
  25. copa_train/few_shot_1/meta-llama_Llama-2-7b-chat-hf/last_token_relative_depth_100/accuracy_blocks.25.hook_resid_post.json +1 -0
  26. copa_train/few_shot_1/meta-llama_Llama-2-7b-chat-hf/last_token_relative_depth_100/accuracy_blocks.26.hook_mlp_out.json +1 -0
  27. copa_train/few_shot_1/meta-llama_Llama-2-7b-chat-hf/last_token_relative_depth_100/accuracy_blocks.27.hook_mlp_out.json +1 -0
  28. copa_train/few_shot_1/meta-llama_Llama-2-7b-chat-hf/last_token_relative_depth_100/accuracy_blocks.3.hook_mlp_out.json +1 -0
  29. copa_train/few_shot_1/meta-llama_Llama-2-7b-chat-hf/last_token_relative_depth_100/accuracy_blocks.3.hook_resid_post.json +1 -0
  30. copa_train/few_shot_1/meta-llama_Llama-2-7b-chat-hf/last_token_relative_depth_100/accuracy_blocks.4.hook_mlp_out.json +1 -0
  31. copa_train/few_shot_1/meta-llama_Llama-2-7b-chat-hf/last_token_relative_depth_100/accuracy_blocks.5.hook_mlp_out.json +1 -0
  32. copa_train/few_shot_1/meta-llama_Llama-2-7b-chat-hf/last_token_relative_depth_100/accuracy_blocks.6.hook_mlp_out.json +1 -0
  33. copa_train/few_shot_1/meta-llama_Llama-2-7b-chat-hf/last_token_relative_depth_100/accuracy_blocks.6.hook_resid_post.json +1 -0
  34. copa_train/few_shot_1/meta-llama_Llama-2-7b-chat-hf/last_token_relative_depth_100/accuracy_blocks.7.hook_mlp_out.json +1 -0
  35. copa_train/few_shot_1/meta-llama_Llama-2-7b-chat-hf/last_token_relative_depth_100/accuracy_blocks.7.hook_resid_post.json +1 -0
  36. copa_train/few_shot_1/meta-llama_Llama-2-7b-chat-hf/last_token_relative_depth_100/accuracy_blocks.8.hook_resid_post.json +1 -0
  37. copa_train/few_shot_1/meta-llama_Llama-2-7b-chat-hf/last_token_relative_depth_100/accuracy_blocks.9.hook_resid_post.json +1 -0
  38. copa_train/few_shot_1/meta-llama_Llama-2-7b-chat-hf/last_token_relative_depth_100/accuracy_ln_final.hook_normalized.json +1 -0
  39. copa_train/few_shot_1/meta-llama_Llama-2-7b-chat-hf/last_token_relative_depth_100/accuracy_output.json +1 -0
  40. copa_train/few_shot_1/meta-llama_Llama-2-7b-chat-hf/last_token_relative_depth_100/intrinsic/mle_local_dims/blocks.0.hook_mlp_out.csv +0 -0
  41. copa_train/few_shot_1/meta-llama_Llama-2-7b-chat-hf/last_token_relative_depth_100/intrinsic/mle_local_dims/blocks.0.hook_resid_post.csv +0 -0
  42. copa_train/few_shot_1/meta-llama_Llama-2-7b-chat-hf/last_token_relative_depth_100/intrinsic/mle_local_dims/blocks.1.hook_resid_post.csv +0 -0
  43. copa_train/few_shot_1/meta-llama_Llama-2-7b-chat-hf/last_token_relative_depth_100/intrinsic/mle_local_dims/blocks.12.hook_mlp_out.csv +0 -0
  44. copa_train/few_shot_1/meta-llama_Llama-2-7b-chat-hf/last_token_relative_depth_100/intrinsic/mle_local_dims/blocks.13.hook_mlp_out.csv +0 -0
  45. copa_train/few_shot_1/meta-llama_Llama-2-7b-chat-hf/last_token_relative_depth_100/intrinsic/mle_local_dims/blocks.16.hook_mlp_out.csv +0 -0
  46. copa_train/few_shot_1/meta-llama_Llama-2-7b-chat-hf/last_token_relative_depth_100/intrinsic/mle_local_dims/blocks.18.hook_mlp_out.csv +0 -0
  47. copa_train/few_shot_1/meta-llama_Llama-2-7b-chat-hf/last_token_relative_depth_100/intrinsic/mle_local_dims/blocks.19.hook_mlp_out.csv +0 -0
  48. copa_train/few_shot_1/meta-llama_Llama-2-7b-chat-hf/last_token_relative_depth_100/intrinsic/mle_local_dims/blocks.2.hook_resid_post.csv +0 -0
  49. copa_train/few_shot_1/meta-llama_Llama-2-7b-chat-hf/last_token_relative_depth_100/intrinsic/mle_local_dims/blocks.20.hook_resid_post.csv +0 -0
  50. copa_train/few_shot_1/meta-llama_Llama-2-7b-chat-hf/last_token_relative_depth_100/intrinsic/mle_local_dims/blocks.21.hook_resid_post.csv +0 -0
copa_train/few_shot_1/meta-llama_Llama-2-7b-chat-hf/last_token_relative_depth_100/accuracy_blocks.0.hook_mlp_out.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"total_per_class": {"0": 489, "1": 511}, "correct_per_class": {"0": 489, "1": 0}, "net_accuracy": 0.489}
copa_train/few_shot_1/meta-llama_Llama-2-7b-chat-hf/last_token_relative_depth_100/accuracy_blocks.11.hook_mlp_out.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"total_per_class": {"0": 489, "1": 511}, "correct_per_class": {"0": 0, "1": 511}, "net_accuracy": 0.511}
copa_train/few_shot_1/meta-llama_Llama-2-7b-chat-hf/last_token_relative_depth_100/accuracy_blocks.12.hook_mlp_out.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"total_per_class": {"0": 489, "1": 511}, "correct_per_class": {"0": 67, "1": 467}, "net_accuracy": 0.534}
copa_train/few_shot_1/meta-llama_Llama-2-7b-chat-hf/last_token_relative_depth_100/accuracy_blocks.12.hook_resid_post.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"total_per_class": {"0": 489, "1": 511}, "correct_per_class": {"0": 0, "1": 511}, "net_accuracy": 0.511}
copa_train/few_shot_1/meta-llama_Llama-2-7b-chat-hf/last_token_relative_depth_100/accuracy_blocks.13.hook_mlp_out.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"total_per_class": {"0": 489, "1": 511}, "correct_per_class": {"0": 0, "1": 509}, "net_accuracy": 0.509}
copa_train/few_shot_1/meta-llama_Llama-2-7b-chat-hf/last_token_relative_depth_100/accuracy_blocks.13.hook_resid_post.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"total_per_class": {"0": 489, "1": 511}, "correct_per_class": {"0": 0, "1": 511}, "net_accuracy": 0.511}
copa_train/few_shot_1/meta-llama_Llama-2-7b-chat-hf/last_token_relative_depth_100/accuracy_blocks.14.hook_mlp_out.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"total_per_class": {"0": 489, "1": 511}, "correct_per_class": {"0": 451, "1": 40}, "net_accuracy": 0.491}
copa_train/few_shot_1/meta-llama_Llama-2-7b-chat-hf/last_token_relative_depth_100/accuracy_blocks.15.hook_mlp_out.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"total_per_class": {"0": 489, "1": 511}, "correct_per_class": {"0": 25, "1": 463}, "net_accuracy": 0.488}
copa_train/few_shot_1/meta-llama_Llama-2-7b-chat-hf/last_token_relative_depth_100/accuracy_blocks.16.hook_mlp_out.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"total_per_class": {"0": 489, "1": 511}, "correct_per_class": {"0": 489, "1": 0}, "net_accuracy": 0.489}
copa_train/few_shot_1/meta-llama_Llama-2-7b-chat-hf/last_token_relative_depth_100/accuracy_blocks.16.hook_resid_post.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"total_per_class": {"0": 489, "1": 511}, "correct_per_class": {"0": 14, "1": 511}, "net_accuracy": 0.525}
copa_train/few_shot_1/meta-llama_Llama-2-7b-chat-hf/last_token_relative_depth_100/accuracy_blocks.17.hook_mlp_out.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"total_per_class": {"0": 489, "1": 511}, "correct_per_class": {"0": 484, "1": 8}, "net_accuracy": 0.492}
copa_train/few_shot_1/meta-llama_Llama-2-7b-chat-hf/last_token_relative_depth_100/accuracy_blocks.17.hook_resid_post.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"total_per_class": {"0": 489, "1": 511}, "correct_per_class": {"0": 53, "1": 509}, "net_accuracy": 0.562}
copa_train/few_shot_1/meta-llama_Llama-2-7b-chat-hf/last_token_relative_depth_100/accuracy_blocks.18.hook_resid_post.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"total_per_class": {"0": 489, "1": 511}, "correct_per_class": {"0": 271, "1": 452}, "net_accuracy": 0.723}
copa_train/few_shot_1/meta-llama_Llama-2-7b-chat-hf/last_token_relative_depth_100/accuracy_blocks.19.hook_resid_post.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"total_per_class": {"0": 489, "1": 511}, "correct_per_class": {"0": 278, "1": 452}, "net_accuracy": 0.73}
copa_train/few_shot_1/meta-llama_Llama-2-7b-chat-hf/last_token_relative_depth_100/accuracy_blocks.2.hook_mlp_out.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"total_per_class": {"0": 489, "1": 511}, "correct_per_class": {"0": 0, "1": 511}, "net_accuracy": 0.511}
copa_train/few_shot_1/meta-llama_Llama-2-7b-chat-hf/last_token_relative_depth_100/accuracy_blocks.2.hook_resid_post.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"total_per_class": {"0": 489, "1": 511}, "correct_per_class": {"0": 489, "1": 0}, "net_accuracy": 0.489}
copa_train/few_shot_1/meta-llama_Llama-2-7b-chat-hf/last_token_relative_depth_100/accuracy_blocks.20.hook_mlp_out.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"total_per_class": {"0": 489, "1": 511}, "correct_per_class": {"0": 118, "1": 280}, "net_accuracy": 0.398}
copa_train/few_shot_1/meta-llama_Llama-2-7b-chat-hf/last_token_relative_depth_100/accuracy_blocks.21.hook_mlp_out.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"total_per_class": {"0": 489, "1": 511}, "correct_per_class": {"0": 484, "1": 4}, "net_accuracy": 0.488}
copa_train/few_shot_1/meta-llama_Llama-2-7b-chat-hf/last_token_relative_depth_100/accuracy_blocks.21.hook_resid_post.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"total_per_class": {"0": 489, "1": 511}, "correct_per_class": {"0": 162, "1": 498}, "net_accuracy": 0.66}
copa_train/few_shot_1/meta-llama_Llama-2-7b-chat-hf/last_token_relative_depth_100/accuracy_blocks.22.hook_mlp_out.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"total_per_class": {"0": 489, "1": 511}, "correct_per_class": {"0": 0, "1": 511}, "net_accuracy": 0.511}
copa_train/few_shot_1/meta-llama_Llama-2-7b-chat-hf/last_token_relative_depth_100/accuracy_blocks.23.hook_mlp_out.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"total_per_class": {"0": 489, "1": 511}, "correct_per_class": {"0": 455, "1": 50}, "net_accuracy": 0.505}
copa_train/few_shot_1/meta-llama_Llama-2-7b-chat-hf/last_token_relative_depth_100/accuracy_blocks.24.hook_mlp_out.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"total_per_class": {"0": 489, "1": 511}, "correct_per_class": {"0": 482, "1": 59}, "net_accuracy": 0.541}
copa_train/few_shot_1/meta-llama_Llama-2-7b-chat-hf/last_token_relative_depth_100/accuracy_blocks.24.hook_resid_post.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"total_per_class": {"0": 489, "1": 511}, "correct_per_class": {"0": 287, "1": 448}, "net_accuracy": 0.735}
copa_train/few_shot_1/meta-llama_Llama-2-7b-chat-hf/last_token_relative_depth_100/accuracy_blocks.25.hook_mlp_out.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"total_per_class": {"0": 489, "1": 511}, "correct_per_class": {"0": 1, "1": 508}, "net_accuracy": 0.509}
copa_train/few_shot_1/meta-llama_Llama-2-7b-chat-hf/last_token_relative_depth_100/accuracy_blocks.25.hook_resid_post.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"total_per_class": {"0": 489, "1": 511}, "correct_per_class": {"0": 237, "1": 468}, "net_accuracy": 0.705}
copa_train/few_shot_1/meta-llama_Llama-2-7b-chat-hf/last_token_relative_depth_100/accuracy_blocks.26.hook_mlp_out.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"total_per_class": {"0": 489, "1": 511}, "correct_per_class": {"0": 361, "1": 268}, "net_accuracy": 0.629}
copa_train/few_shot_1/meta-llama_Llama-2-7b-chat-hf/last_token_relative_depth_100/accuracy_blocks.27.hook_mlp_out.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"total_per_class": {"0": 489, "1": 511}, "correct_per_class": {"0": 281, "1": 85}, "net_accuracy": 0.366}
copa_train/few_shot_1/meta-llama_Llama-2-7b-chat-hf/last_token_relative_depth_100/accuracy_blocks.3.hook_mlp_out.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"total_per_class": {"0": 489, "1": 511}, "correct_per_class": {"0": 489, "1": 0}, "net_accuracy": 0.489}
copa_train/few_shot_1/meta-llama_Llama-2-7b-chat-hf/last_token_relative_depth_100/accuracy_blocks.3.hook_resid_post.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"total_per_class": {"0": 489, "1": 511}, "correct_per_class": {"0": 489, "1": 0}, "net_accuracy": 0.489}
copa_train/few_shot_1/meta-llama_Llama-2-7b-chat-hf/last_token_relative_depth_100/accuracy_blocks.4.hook_mlp_out.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"total_per_class": {"0": 489, "1": 511}, "correct_per_class": {"0": 0, "1": 510}, "net_accuracy": 0.51}
copa_train/few_shot_1/meta-llama_Llama-2-7b-chat-hf/last_token_relative_depth_100/accuracy_blocks.5.hook_mlp_out.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"total_per_class": {"0": 489, "1": 511}, "correct_per_class": {"0": 0, "1": 511}, "net_accuracy": 0.511}
copa_train/few_shot_1/meta-llama_Llama-2-7b-chat-hf/last_token_relative_depth_100/accuracy_blocks.6.hook_mlp_out.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"total_per_class": {"0": 489, "1": 511}, "correct_per_class": {"0": 0, "1": 511}, "net_accuracy": 0.511}
copa_train/few_shot_1/meta-llama_Llama-2-7b-chat-hf/last_token_relative_depth_100/accuracy_blocks.6.hook_resid_post.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"total_per_class": {"0": 489, "1": 511}, "correct_per_class": {"0": 30, "1": 478}, "net_accuracy": 0.508}
copa_train/few_shot_1/meta-llama_Llama-2-7b-chat-hf/last_token_relative_depth_100/accuracy_blocks.7.hook_mlp_out.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"total_per_class": {"0": 489, "1": 511}, "correct_per_class": {"0": 0, "1": 511}, "net_accuracy": 0.511}
copa_train/few_shot_1/meta-llama_Llama-2-7b-chat-hf/last_token_relative_depth_100/accuracy_blocks.7.hook_resid_post.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"total_per_class": {"0": 489, "1": 511}, "correct_per_class": {"0": 0, "1": 511}, "net_accuracy": 0.511}
copa_train/few_shot_1/meta-llama_Llama-2-7b-chat-hf/last_token_relative_depth_100/accuracy_blocks.8.hook_resid_post.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"total_per_class": {"0": 489, "1": 511}, "correct_per_class": {"0": 0, "1": 511}, "net_accuracy": 0.511}
copa_train/few_shot_1/meta-llama_Llama-2-7b-chat-hf/last_token_relative_depth_100/accuracy_blocks.9.hook_resid_post.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"total_per_class": {"0": 489, "1": 511}, "correct_per_class": {"0": 0, "1": 511}, "net_accuracy": 0.511}
copa_train/few_shot_1/meta-llama_Llama-2-7b-chat-hf/last_token_relative_depth_100/accuracy_ln_final.hook_normalized.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"total_per_class": {"0": 489, "1": 511}, "correct_per_class": {"0": 229, "1": 468}, "net_accuracy": 0.697}
copa_train/few_shot_1/meta-llama_Llama-2-7b-chat-hf/last_token_relative_depth_100/accuracy_output.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"total_per_class": {"0": 489, "1": 511}, "correct_per_class": {"0": 229, "1": 468}, "net_accuracy": 0.697}
copa_train/few_shot_1/meta-llama_Llama-2-7b-chat-hf/last_token_relative_depth_100/intrinsic/mle_local_dims/blocks.0.hook_mlp_out.csv ADDED
The diff for this file is too large to render. See raw diff
 
copa_train/few_shot_1/meta-llama_Llama-2-7b-chat-hf/last_token_relative_depth_100/intrinsic/mle_local_dims/blocks.0.hook_resid_post.csv ADDED
The diff for this file is too large to render. See raw diff
 
copa_train/few_shot_1/meta-llama_Llama-2-7b-chat-hf/last_token_relative_depth_100/intrinsic/mle_local_dims/blocks.1.hook_resid_post.csv ADDED
The diff for this file is too large to render. See raw diff
 
copa_train/few_shot_1/meta-llama_Llama-2-7b-chat-hf/last_token_relative_depth_100/intrinsic/mle_local_dims/blocks.12.hook_mlp_out.csv ADDED
The diff for this file is too large to render. See raw diff
 
copa_train/few_shot_1/meta-llama_Llama-2-7b-chat-hf/last_token_relative_depth_100/intrinsic/mle_local_dims/blocks.13.hook_mlp_out.csv ADDED
The diff for this file is too large to render. See raw diff
 
copa_train/few_shot_1/meta-llama_Llama-2-7b-chat-hf/last_token_relative_depth_100/intrinsic/mle_local_dims/blocks.16.hook_mlp_out.csv ADDED
The diff for this file is too large to render. See raw diff
 
copa_train/few_shot_1/meta-llama_Llama-2-7b-chat-hf/last_token_relative_depth_100/intrinsic/mle_local_dims/blocks.18.hook_mlp_out.csv ADDED
The diff for this file is too large to render. See raw diff
 
copa_train/few_shot_1/meta-llama_Llama-2-7b-chat-hf/last_token_relative_depth_100/intrinsic/mle_local_dims/blocks.19.hook_mlp_out.csv ADDED
The diff for this file is too large to render. See raw diff
 
copa_train/few_shot_1/meta-llama_Llama-2-7b-chat-hf/last_token_relative_depth_100/intrinsic/mle_local_dims/blocks.2.hook_resid_post.csv ADDED
The diff for this file is too large to render. See raw diff
 
copa_train/few_shot_1/meta-llama_Llama-2-7b-chat-hf/last_token_relative_depth_100/intrinsic/mle_local_dims/blocks.20.hook_resid_post.csv ADDED
The diff for this file is too large to render. See raw diff
 
copa_train/few_shot_1/meta-llama_Llama-2-7b-chat-hf/last_token_relative_depth_100/intrinsic/mle_local_dims/blocks.21.hook_resid_post.csv ADDED
The diff for this file is too large to render. See raw diff