iarcuschin
commited on
Commit
•
2aba647
1
Parent(s):
43c361b
Update metadata
Browse files- benchmark_cases_metadata.csv +19 -19
- benchmark_cases_metadata.parquet +2 -2
- benchmark_metadata.json +474 -314
- benchmark_metadata_croissant.json +150 -135
benchmark_cases_metadata.csv
CHANGED
@@ -1,19 +1,19 @@
|
|
1 |
-
case_id,url,task_description,max_seq_len,min_seq_len,transformer_cfg_file_url,training_args_file_url,weights_file_url,circuit_file_url,training_args.atol,training_args.lr,training_args.use_single_loss,training_args.iit_weight,training_args.behavior_weight,training_args.strict_weight,training_args.epochs,training_args.act_fn,training_args.clip_grad_norm,training_args.lr_scheduler,transformer_cfg.n_layers,transformer_cfg.d_model,transformer_cfg.n_ctx,transformer_cfg.d_head,transformer_cfg.model_name,transformer_cfg.n_heads,transformer_cfg.d_mlp,transformer_cfg.act_fn,transformer_cfg.d_vocab,transformer_cfg.eps,transformer_cfg.use_attn_result,transformer_cfg.use_attn_scale,transformer_cfg.use_split_qkv_input,transformer_cfg.use_hook_mlp_in,transformer_cfg.use_attn_in,transformer_cfg.use_local_attn,transformer_cfg.original_architecture,transformer_cfg.from_checkpoint,transformer_cfg.
|
2 |
-
11,https://huggingface.co/cybershiptrooper/InterpBench/tree/main/11,Counts the number of words in a sequence based on their length.,10,4,https://huggingface.co/cybershiptrooper/InterpBench/blob/main/11/
|
3 |
-
13,https://huggingface.co/cybershiptrooper/InterpBench/tree/main/13,"Analyzes the trend (increasing, decreasing, constant) of numeric tokens.",10,4,https://huggingface.co/cybershiptrooper/InterpBench/blob/main/13/
|
4 |
-
18,https://huggingface.co/cybershiptrooper/InterpBench/tree/main/18,"Classify each token based on its frequency as 'rare', 'common', or 'frequent'.",10,4,https://huggingface.co/cybershiptrooper/InterpBench/blob/main/18/
|
5 |
-
19,https://huggingface.co/cybershiptrooper/InterpBench/tree/main/19,Removes consecutive duplicate tokens from a sequence.,15,4,https://huggingface.co/cybershiptrooper/InterpBench/blob/main/19/
|
6 |
-
20,https://huggingface.co/cybershiptrooper/InterpBench/tree/main/20,Detect spam messages based on appearance of spam keywords.,10,4,https://huggingface.co/cybershiptrooper/InterpBench/blob/main/20/
|
7 |
-
21,https://huggingface.co/cybershiptrooper/InterpBench/tree/main/21,Extract unique tokens from a string,10,4,https://huggingface.co/cybershiptrooper/InterpBench/blob/main/21/
|
8 |
-
|
9 |
-
|
10 |
-
|
11 |
-
|
12 |
-
|
13 |
-
|
14 |
-
|
15 |
-
|
16 |
-
4,https://huggingface.co/cybershiptrooper/InterpBench/tree/main/4,Return fraction of previous open tokens minus the fraction of close tokens.,10,4,https://huggingface.co/cybershiptrooper/InterpBench/blob/main/4/
|
17 |
-
8,https://huggingface.co/cybershiptrooper/InterpBench/tree/main/8,Fills gaps between tokens with a specified filler.,10,4,https://huggingface.co/cybershiptrooper/InterpBench/blob/main/8/
|
18 |
-
ioi,https://huggingface.co/cybershiptrooper/InterpBench/tree/main/ioi,Indirect object identification,16,16
|
19 |
-
ioi_next_token,https://huggingface.co/cybershiptrooper/InterpBench/tree/main/ioi_next_token,Indirect object identification,16,16
|
|
|
1 |
+
case_id,url,task_description,max_seq_len,min_seq_len,transformer_cfg_file_url,training_args_file_url,weights_file_url,circuit_file_url,training_args.atol,training_args.lr,training_args.use_single_loss,training_args.iit_weight,training_args.behavior_weight,training_args.strict_weight,training_args.epochs,training_args.act_fn,training_args.clip_grad_norm,training_args.lr_scheduler,transformer_cfg.n_layers,transformer_cfg.d_model,transformer_cfg.n_ctx,transformer_cfg.d_head,transformer_cfg.model_name,transformer_cfg.n_heads,transformer_cfg.d_mlp,transformer_cfg.act_fn,transformer_cfg.d_vocab,transformer_cfg.eps,transformer_cfg.use_attn_result,transformer_cfg.use_attn_scale,transformer_cfg.use_split_qkv_input,transformer_cfg.use_hook_mlp_in,transformer_cfg.use_attn_in,transformer_cfg.use_local_attn,transformer_cfg.original_architecture,transformer_cfg.from_checkpoint,transformer_cfg.tokenizer_name,transformer_cfg.init_mode,transformer_cfg.normalization_type,transformer_cfg.n_devices,transformer_cfg.attention_dir,transformer_cfg.attn_only,transformer_cfg.seed,transformer_cfg.initializer_range,transformer_cfg.init_weights,transformer_cfg.scale_attn_by_inverse_layer_idx,transformer_cfg.positional_embedding_type,transformer_cfg.final_rms,transformer_cfg.d_vocab_out,transformer_cfg.parallel_attn_mlp,transformer_cfg.n_params,transformer_cfg.use_hook_tokens,transformer_cfg.gated_mlp,transformer_cfg.default_prepend_bos,transformer_cfg.dtype,transformer_cfg.tokenizer_prepends_bos,transformer_cfg.post_embedding_ln,transformer_cfg.rotary_base,transformer_cfg.trust_remote_code,transformer_cfg.rotary_adjacent_pairs,training_args.model_pair,training_args.next_token,training_args.non_ioi_thresh,training_args.use_per_token_check,training_args.batch_size,training_args.num_workers,training_args.early_stop,training_args.scheduler_val_metric,training_args.scheduler_mode
|
2 |
+
11,https://huggingface.co/cybershiptrooper/InterpBench/tree/main/11,Counts the number of words in a sequence based on their length.,10,4,https://huggingface.co/cybershiptrooper/InterpBench/blob/main/11/ll_model_cfg.pkl,https://huggingface.co/cybershiptrooper/InterpBench/blob/main/11/meta.json,https://huggingface.co/cybershiptrooper/InterpBench/blob/main/11/ll_model.pth,https://huggingface.co/cybershiptrooper/InterpBench/blob/main/11/edges.pkl,0.05,0.01,False,1.0,1.0,0.4,500.0,gelu,1.0,,2,12,10,3,custom,4,48,gelu,10,1e-05,True,True,True,True,False,False,,False,,gpt2,,1,causal,False,0.0,0.1460593486680443,True,False,standard,False,5,False,3456,False,False,True,torch.float32,False,False,10000,False,False,,True,,True,,,True,,
|
3 |
+
13,https://huggingface.co/cybershiptrooper/InterpBench/tree/main/13,"Analyzes the trend (increasing, decreasing, constant) of numeric tokens.",10,4,https://huggingface.co/cybershiptrooper/InterpBench/blob/main/13/ll_model_cfg.pkl,https://huggingface.co/cybershiptrooper/InterpBench/blob/main/13/meta.json,https://huggingface.co/cybershiptrooper/InterpBench/blob/main/13/ll_model.pth,https://huggingface.co/cybershiptrooper/InterpBench/blob/main/13/edges.pkl,0.05,0.01,False,1.0,1.0,0.4,500.0,gelu,1.0,,2,20,10,5,custom,4,80,gelu,5,1e-05,True,True,True,True,False,False,,False,,gpt2,,1,bidirectional,False,0.0,0.1460593486680443,True,False,standard,False,3,False,9600,False,False,True,torch.float32,False,False,10000,False,False,,True,,True,,,True,,
|
4 |
+
18,https://huggingface.co/cybershiptrooper/InterpBench/tree/main/18,"Classify each token based on its frequency as 'rare', 'common', or 'frequent'.",10,4,https://huggingface.co/cybershiptrooper/InterpBench/blob/main/18/ll_model_cfg.pkl,https://huggingface.co/cybershiptrooper/InterpBench/blob/main/18/meta.json,https://huggingface.co/cybershiptrooper/InterpBench/blob/main/18/ll_model.pth,https://huggingface.co/cybershiptrooper/InterpBench/blob/main/18/edges.pkl,0.05,0.001,False,1.0,1.0,1.0,2000.0,gelu,0.1,,2,26,10,6,custom,4,104,gelu,7,1e-05,True,True,True,True,False,False,,False,,gpt2,,1,bidirectional,False,0.0,0.12344267996967354,True,False,standard,False,3,False,15808,False,False,True,torch.float32,False,False,10000,False,False,strict,True,,True,,,True,,
|
5 |
+
19,https://huggingface.co/cybershiptrooper/InterpBench/tree/main/19,Removes consecutive duplicate tokens from a sequence.,15,4,https://huggingface.co/cybershiptrooper/InterpBench/blob/main/19/ll_model_cfg.pkl,https://huggingface.co/cybershiptrooper/InterpBench/blob/main/19/meta.json,https://huggingface.co/cybershiptrooper/InterpBench/blob/main/19/ll_model.pth,https://huggingface.co/cybershiptrooper/InterpBench/blob/main/19/edges.pkl,0.05,0.001,False,1.0,1.0,0.4,2000.0,gelu,0.1,,2,32,15,8,custom,4,128,gelu,5,1e-05,True,True,True,True,False,False,,False,,gpt2,,1,causal,False,0.0,0.15689290811054724,True,False,standard,False,3,False,24576,False,False,True,torch.float32,False,False,10000,False,False,,True,,True,,,True,,
|
6 |
+
20,https://huggingface.co/cybershiptrooper/InterpBench/tree/main/20,Detect spam messages based on appearance of spam keywords.,10,4,https://huggingface.co/cybershiptrooper/InterpBench/blob/main/20/ll_model_cfg.pkl,https://huggingface.co/cybershiptrooper/InterpBench/blob/main/20/meta.json,https://huggingface.co/cybershiptrooper/InterpBench/blob/main/20/ll_model.pth,https://huggingface.co/cybershiptrooper/InterpBench/blob/main/20/edges.pkl,0.05,0.01,False,1.0,1.0,0.4,2000.0,gelu,0.1,,2,13,10,3,custom,4,52,gelu,14,1e-05,True,True,True,True,False,False,,False,,gpt2,,1,causal,False,0.0,0.16,True,False,standard,False,2,False,3952,False,False,True,torch.float32,False,False,10000,False,False,strict,True,,True,,,True,,
|
7 |
+
21,https://huggingface.co/cybershiptrooper/InterpBench/tree/main/21,Extract unique tokens from a string,10,4,https://huggingface.co/cybershiptrooper/InterpBench/blob/main/21/ll_model_cfg.pkl,https://huggingface.co/cybershiptrooper/InterpBench/blob/main/21/meta.json,https://huggingface.co/cybershiptrooper/InterpBench/blob/main/21/ll_model.pth,https://huggingface.co/cybershiptrooper/InterpBench/blob/main/21/edges.pkl,0.05,0.0005,False,1.0,1.0,0.5,2000.0,gelu,0.1,,4,50,10,12,custom,4,200,gelu,5,1e-05,True,True,True,True,False,False,,False,,gpt2,,1,causal,False,0.0,0.09847319278346618,True,False,standard,False,3,False,118400,False,False,True,torch.float32,False,False,10000,False,False,strict,True,,True,,,True,,
|
8 |
+
26,https://huggingface.co/cybershiptrooper/InterpBench/tree/main/26,Creates a cascading effect by repeating each token in sequence incrementally.,10,4,https://huggingface.co/cybershiptrooper/InterpBench/blob/main/26/ll_model_cfg.pkl,https://huggingface.co/cybershiptrooper/InterpBench/blob/main/26/meta.json,https://huggingface.co/cybershiptrooper/InterpBench/blob/main/26/ll_model.pth,https://huggingface.co/cybershiptrooper/InterpBench/blob/main/26/edges.pkl,0.05,0.01,False,1.0,1.0,0.4,2000.0,gelu,0.1,,2,21,10,5,custom,4,84,gelu,5,1e-05,True,True,True,True,False,False,,False,,gpt2,,1,causal,False,0.0,0.12344267996967354,True,False,standard,False,27,False,10416,False,False,True,torch.float32,False,False,10000,False,False,strict,True,,True,,,True,,
|
9 |
+
29,https://huggingface.co/cybershiptrooper/InterpBench/tree/main/29,Creates abbreviations for each token in the sequence.,10,4,https://huggingface.co/cybershiptrooper/InterpBench/blob/main/29/ll_model_cfg.pkl,https://huggingface.co/cybershiptrooper/InterpBench/blob/main/29/meta.json,https://huggingface.co/cybershiptrooper/InterpBench/blob/main/29/ll_model.pth,https://huggingface.co/cybershiptrooper/InterpBench/blob/main/29/edges.pkl,0.05,0.01,False,1.0,1.0,0.4,2000.0,gelu,0.1,,2,13,10,3,custom,4,52,gelu,10,1e-05,True,True,True,True,False,False,,False,,gpt2,,1,causal,False,0.0,0.1539600717839002,True,False,standard,False,8,False,3952,False,False,True,torch.float32,False,False,10000,False,False,strict,True,,True,,,True,,
|
10 |
+
3,https://huggingface.co/cybershiptrooper/InterpBench/tree/main/3,Returns the fraction of 'x' in the input up to the i-th position for all i.,5,4,https://huggingface.co/cybershiptrooper/InterpBench/blob/main/3/ll_model_cfg.pkl,https://huggingface.co/cybershiptrooper/InterpBench/blob/main/3/meta.json,https://huggingface.co/cybershiptrooper/InterpBench/blob/main/3/ll_model.pth,https://huggingface.co/cybershiptrooper/InterpBench/blob/main/3/edges.pkl,0.05,0.001,False,1.0,1.0,10.0,2000.0,gelu,0.1,,2,12,5,3,custom,4,48,gelu,6,1e-05,True,True,True,True,False,False,,False,,gpt2,,1,causal,False,0.0,0.22188007849009167,True,False,standard,False,1,False,3456,False,False,True,torch.float32,False,False,10000,False,False,,True,,True,,,True,,
|
11 |
+
33,https://huggingface.co/cybershiptrooper/InterpBench/tree/main/33,Checks if each token's length is odd or even.,10,4,https://huggingface.co/cybershiptrooper/InterpBench/blob/main/33/ll_model_cfg.pkl,https://huggingface.co/cybershiptrooper/InterpBench/blob/main/33/meta.json,https://huggingface.co/cybershiptrooper/InterpBench/blob/main/33/ll_model.pth,https://huggingface.co/cybershiptrooper/InterpBench/blob/main/33/edges.pkl,0.05,0.001,False,1.0,1.0,0.4,2000.0,gelu,0.1,,2,4,10,1,custom,4,16,gelu,10,1e-05,True,True,True,True,False,False,,False,,gpt2,,1,causal,False,0.0,0.17457431218879393,True,False,standard,False,2,False,384,False,False,True,torch.float32,False,False,10000,False,False,,True,,True,,,True,,
|
12 |
+
34,https://huggingface.co/cybershiptrooper/InterpBench/tree/main/34,Calculate the ratio of vowels to consonants in each word.,10,4,https://huggingface.co/cybershiptrooper/InterpBench/blob/main/34/ll_model_cfg.pkl,https://huggingface.co/cybershiptrooper/InterpBench/blob/main/34/meta.json,https://huggingface.co/cybershiptrooper/InterpBench/blob/main/34/ll_model.pth,https://huggingface.co/cybershiptrooper/InterpBench/blob/main/34/edges.pkl,0.05,0.01,False,1.0,1.0,1.0,2000.0,gelu,0.1,,2,16,10,4,custom,4,64,gelu,10,1e-05,True,True,True,True,False,False,,False,,gpt2,,1,causal,False,0.0,0.16329931618554522,True,False,standard,False,5,False,6144,False,False,True,torch.float32,False,False,10000,False,False,strict,True,,True,,,True,,
|
13 |
+
35,https://huggingface.co/cybershiptrooper/InterpBench/tree/main/35,Alternates capitalization of each character in words.,10,4,https://huggingface.co/cybershiptrooper/InterpBench/blob/main/35/ll_model_cfg.pkl,https://huggingface.co/cybershiptrooper/InterpBench/blob/main/35/meta.json,https://huggingface.co/cybershiptrooper/InterpBench/blob/main/35/ll_model.pth,https://huggingface.co/cybershiptrooper/InterpBench/blob/main/35/edges.pkl,0.05,0.01,False,1.0,1.0,1.0,2000.0,gelu,0.1,,2,9,10,2,custom,4,36,gelu,10,1e-05,True,True,True,True,False,False,,False,,gpt2,,1,causal,False,0.0,0.1539600717839002,True,False,standard,False,8,False,1872,False,False,True,torch.float32,False,False,10000,False,False,strict,True,,True,,,True,,
|
14 |
+
36,https://huggingface.co/cybershiptrooper/InterpBench/tree/main/36,"Classifies each token as 'positive', 'negative', or 'neutral' based on emojis.",10,4,https://huggingface.co/cybershiptrooper/InterpBench/blob/main/36/ll_model_cfg.pkl,https://huggingface.co/cybershiptrooper/InterpBench/blob/main/36/meta.json,https://huggingface.co/cybershiptrooper/InterpBench/blob/main/36/ll_model.pth,https://huggingface.co/cybershiptrooper/InterpBench/blob/main/36/edges.pkl,0.05,0.01,False,1.0,1.0,1.0,2000.0,gelu,0.1,,2,6,10,1,custom,4,24,gelu,5,1e-05,True,True,True,True,False,False,,False,,gpt2,,1,causal,False,0.0,0.19402850002906638,True,False,standard,False,3,False,768,False,False,True,torch.float32,False,False,10000,False,False,strict,True,,True,,,True,,
|
15 |
+
37,https://huggingface.co/cybershiptrooper/InterpBench/tree/main/37,Reverses each word in the sequence except for specified exclusions.,10,4,https://huggingface.co/cybershiptrooper/InterpBench/blob/main/37/ll_model_cfg.pkl,https://huggingface.co/cybershiptrooper/InterpBench/blob/main/37/meta.json,https://huggingface.co/cybershiptrooper/InterpBench/blob/main/37/ll_model.pth,https://huggingface.co/cybershiptrooper/InterpBench/blob/main/37/edges.pkl,0.05,0.01,False,1.0,1.0,1.0,2000.0,gelu,0.1,,2,12,10,3,custom,4,48,gelu,10,1e-05,True,True,True,True,False,False,,False,,gpt2,,1,causal,False,0.0,0.1539600717839002,True,False,standard,False,8,False,3456,False,False,True,torch.float32,False,False,10000,False,False,strict,True,,True,,,True,,
|
16 |
+
4,https://huggingface.co/cybershiptrooper/InterpBench/tree/main/4,Return fraction of previous open tokens minus the fraction of close tokens.,10,4,https://huggingface.co/cybershiptrooper/InterpBench/blob/main/4/ll_model_cfg.pkl,https://huggingface.co/cybershiptrooper/InterpBench/blob/main/4/meta.json,https://huggingface.co/cybershiptrooper/InterpBench/blob/main/4/ll_model.pth,https://huggingface.co/cybershiptrooper/InterpBench/blob/main/4/edges.pkl,0.05,0.001,False,1.0,1.0,0.4,2000.0,gelu,0.1,,2,20,10,5,custom,4,80,gelu,7,1e-05,True,True,True,True,False,False,,False,,gpt2,,1,causal,False,0.0,0.17056057308448835,True,False,standard,False,1,False,9600,False,False,True,torch.float32,False,False,10000,False,False,,True,,True,,,True,,
|
17 |
+
8,https://huggingface.co/cybershiptrooper/InterpBench/tree/main/8,Fills gaps between tokens with a specified filler.,10,4,https://huggingface.co/cybershiptrooper/InterpBench/blob/main/8/ll_model_cfg.pkl,https://huggingface.co/cybershiptrooper/InterpBench/blob/main/8/meta.json,https://huggingface.co/cybershiptrooper/InterpBench/blob/main/8/ll_model.pth,https://huggingface.co/cybershiptrooper/InterpBench/blob/main/8/edges.pkl,0.05,0.01,False,1.0,1.0,0.4,500.0,gelu,1.0,,2,20,10,5,custom,4,80,gelu,10,1e-05,True,True,True,True,False,False,,False,,gpt2,,1,causal,False,0.0,0.13333333333333333,True,False,standard,False,8,False,9600,False,False,True,torch.float32,False,False,10000,False,False,,True,,True,,,True,,
|
18 |
+
ioi,https://huggingface.co/cybershiptrooper/InterpBench/tree/main/ioi,Indirect object identification,16,16,https://huggingface.co/cybershiptrooper/InterpBench/blob/main/ioi/ll_model_cfg.pkl,https://huggingface.co/cybershiptrooper/InterpBench/blob/main/ioi/meta.json,https://huggingface.co/cybershiptrooper/InterpBench/blob/main/ioi/ll_model.pth,https://huggingface.co/cybershiptrooper/InterpBench/blob/main/ioi/edges.pkl,0.05,0.001,False,1.0,1.0,0.4,,,1.0,,6,64,1024,16,gpt2,4,3072,gelu_new,50257,1e-05,False,True,False,False,False,False,GPT2LMHeadModel,False,gpt2,gpt2,LNPre,1,causal,False,,0.02886751345948129,True,False,standard,False,50257,False,2457600,False,False,True,torch.float32,False,False,10000,False,False,,True,0.65,False,256.0,0.0,True,"val/accuracy,val/IIA",max
|
19 |
+
ioi_next_token,https://huggingface.co/cybershiptrooper/InterpBench/tree/main/ioi_next_token,Indirect object identification,16,16,https://huggingface.co/cybershiptrooper/InterpBench/blob/main/ioi_next_token/ll_model_cfg.pkl,https://huggingface.co/cybershiptrooper/InterpBench/blob/main/ioi_next_token/meta.json,https://huggingface.co/cybershiptrooper/InterpBench/blob/main/ioi_next_token/ll_model.pth,https://huggingface.co/cybershiptrooper/InterpBench/blob/main/ioi_next_token/edges.pkl,0.05,0.001,False,1.0,1.0,0.4,,,1.0,,6,64,1024,16,gpt2,4,3072,gelu_new,50257,1e-05,False,True,False,False,False,False,GPT2LMHeadModel,False,gpt2,gpt2,LNPre,1,causal,False,,0.02886751345948129,True,False,standard,False,50257,False,2457600,False,False,True,torch.float32,False,False,10000,False,False,,True,0.65,False,256.0,0.0,True,"val/accuracy,val/IIA",max
|
benchmark_cases_metadata.parquet
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
-
size
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:568194933b940c4c03457b1c64a8cb074943dfd075ff83f06e84a6376e3a8dcf
|
3 |
+
size 58286
|
benchmark_metadata.json
CHANGED
@@ -27,16 +27,16 @@
|
|
27 |
"url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/11/edges.pkl"
|
28 |
},
|
29 |
{
|
30 |
-
"file_name": "
|
31 |
-
"url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/11/
|
32 |
},
|
33 |
{
|
34 |
-
"file_name": "
|
35 |
-
"url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/11/
|
36 |
},
|
37 |
{
|
38 |
-
"file_name": "
|
39 |
-
"url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/11/
|
40 |
}
|
41 |
],
|
42 |
"transformer_cfg": {
|
@@ -66,7 +66,6 @@
|
|
66 |
"attn_types": null,
|
67 |
"init_mode": "gpt2",
|
68 |
"normalization_type": null,
|
69 |
-
"device": "cpu",
|
70 |
"n_devices": 1,
|
71 |
"attention_dir": "causal",
|
72 |
"attn_only": false,
|
@@ -91,7 +90,7 @@
|
|
91 |
"trust_remote_code": false,
|
92 |
"rotary_adjacent_pairs": false
|
93 |
},
|
94 |
-
"transformer_cfg_file_url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/11/
|
95 |
"training_args": {
|
96 |
"atol": 0.05,
|
97 |
"lr": 0.01,
|
@@ -104,8 +103,8 @@
|
|
104 |
"clip_grad_norm": 1.0,
|
105 |
"lr_scheduler": ""
|
106 |
},
|
107 |
-
"training_args_file_url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/11/
|
108 |
-
"weights_file_url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/11/
|
109 |
"circuit_file_url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/11/edges.pkl"
|
110 |
},
|
111 |
{
|
@@ -125,16 +124,16 @@
|
|
125 |
"url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/13/edges.pkl"
|
126 |
},
|
127 |
{
|
128 |
-
"file_name": "
|
129 |
-
"url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/13/
|
130 |
},
|
131 |
{
|
132 |
-
"file_name": "
|
133 |
-
"url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/13/
|
134 |
},
|
135 |
{
|
136 |
-
"file_name": "
|
137 |
-
"url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/13/
|
138 |
}
|
139 |
],
|
140 |
"transformer_cfg": {
|
@@ -164,7 +163,6 @@
|
|
164 |
"attn_types": null,
|
165 |
"init_mode": "gpt2",
|
166 |
"normalization_type": null,
|
167 |
-
"device": "cpu",
|
168 |
"n_devices": 1,
|
169 |
"attention_dir": "bidirectional",
|
170 |
"attn_only": false,
|
@@ -189,7 +187,7 @@
|
|
189 |
"trust_remote_code": false,
|
190 |
"rotary_adjacent_pairs": false
|
191 |
},
|
192 |
-
"transformer_cfg_file_url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/13/
|
193 |
"training_args": {
|
194 |
"atol": 0.05,
|
195 |
"lr": 0.01,
|
@@ -202,8 +200,8 @@
|
|
202 |
"clip_grad_norm": 1.0,
|
203 |
"lr_scheduler": ""
|
204 |
},
|
205 |
-
"training_args_file_url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/13/
|
206 |
-
"weights_file_url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/13/
|
207 |
"circuit_file_url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/13/edges.pkl"
|
208 |
},
|
209 |
{
|
@@ -225,26 +223,26 @@
|
|
225 |
"url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/18/edges.pkl"
|
226 |
},
|
227 |
{
|
228 |
-
"file_name": "
|
229 |
-
"url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/18/
|
230 |
},
|
231 |
{
|
232 |
-
"file_name": "
|
233 |
-
"url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/18/
|
234 |
},
|
235 |
{
|
236 |
-
"file_name": "
|
237 |
-
"url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/18/
|
238 |
}
|
239 |
],
|
240 |
"transformer_cfg": {
|
241 |
"n_layers": 2,
|
242 |
-
"d_model":
|
243 |
"n_ctx": 10,
|
244 |
-
"d_head":
|
245 |
"model_name": "custom",
|
246 |
"n_heads": 4,
|
247 |
-
"d_mlp":
|
248 |
"act_fn": "gelu",
|
249 |
"d_vocab": 7,
|
250 |
"eps": 1e-05,
|
@@ -264,7 +262,6 @@
|
|
264 |
"attn_types": null,
|
265 |
"init_mode": "gpt2",
|
266 |
"normalization_type": null,
|
267 |
-
"device": "cpu",
|
268 |
"n_devices": 1,
|
269 |
"attention_dir": "bidirectional",
|
270 |
"attn_only": false,
|
@@ -277,7 +274,7 @@
|
|
277 |
"d_vocab_out": 3,
|
278 |
"parallel_attn_mlp": false,
|
279 |
"rotary_dim": null,
|
280 |
-
"n_params":
|
281 |
"use_hook_tokens": false,
|
282 |
"gated_mlp": false,
|
283 |
"default_prepend_bos": true,
|
@@ -289,21 +286,22 @@
|
|
289 |
"trust_remote_code": false,
|
290 |
"rotary_adjacent_pairs": false
|
291 |
},
|
292 |
-
"transformer_cfg_file_url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/18/
|
293 |
"training_args": {
|
294 |
"atol": 0.05,
|
295 |
"lr": 0.001,
|
296 |
"use_single_loss": false,
|
297 |
"iit_weight": 1.0,
|
298 |
"behavior_weight": 1.0,
|
299 |
-
"strict_weight": 0
|
300 |
"epochs": 2000,
|
301 |
"act_fn": "gelu",
|
302 |
"clip_grad_norm": 0.1,
|
303 |
-
"lr_scheduler": ""
|
|
|
304 |
},
|
305 |
-
"training_args_file_url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/18/
|
306 |
-
"weights_file_url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/18/
|
307 |
"circuit_file_url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/18/edges.pkl"
|
308 |
},
|
309 |
{
|
@@ -323,16 +321,16 @@
|
|
323 |
"url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/19/edges.pkl"
|
324 |
},
|
325 |
{
|
326 |
-
"file_name": "
|
327 |
-
"url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/19/
|
328 |
},
|
329 |
{
|
330 |
-
"file_name": "
|
331 |
-
"url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/19/
|
332 |
},
|
333 |
{
|
334 |
-
"file_name": "
|
335 |
-
"url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/19/
|
336 |
}
|
337 |
],
|
338 |
"transformer_cfg": {
|
@@ -362,7 +360,6 @@
|
|
362 |
"attn_types": null,
|
363 |
"init_mode": "gpt2",
|
364 |
"normalization_type": null,
|
365 |
-
"device": "cpu",
|
366 |
"n_devices": 1,
|
367 |
"attention_dir": "causal",
|
368 |
"attn_only": false,
|
@@ -387,7 +384,7 @@
|
|
387 |
"trust_remote_code": false,
|
388 |
"rotary_adjacent_pairs": false
|
389 |
},
|
390 |
-
"transformer_cfg_file_url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/19/
|
391 |
"training_args": {
|
392 |
"atol": 0.05,
|
393 |
"lr": 0.001,
|
@@ -400,8 +397,8 @@
|
|
400 |
"clip_grad_norm": 0.1,
|
401 |
"lr_scheduler": ""
|
402 |
},
|
403 |
-
"training_args_file_url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/19/
|
404 |
-
"weights_file_url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/19/
|
405 |
"circuit_file_url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/19/edges.pkl"
|
406 |
},
|
407 |
{
|
@@ -430,26 +427,26 @@
|
|
430 |
"url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/20/edges.pkl"
|
431 |
},
|
432 |
{
|
433 |
-
"file_name": "
|
434 |
-
"url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/20/
|
435 |
},
|
436 |
{
|
437 |
-
"file_name": "
|
438 |
-
"url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/20/
|
439 |
},
|
440 |
{
|
441 |
-
"file_name": "
|
442 |
-
"url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/20/
|
443 |
}
|
444 |
],
|
445 |
"transformer_cfg": {
|
446 |
"n_layers": 2,
|
447 |
-
"d_model":
|
448 |
"n_ctx": 10,
|
449 |
-
"d_head":
|
450 |
"model_name": "custom",
|
451 |
"n_heads": 4,
|
452 |
-
"d_mlp":
|
453 |
"act_fn": "gelu",
|
454 |
"d_vocab": 14,
|
455 |
"eps": 1e-05,
|
@@ -469,7 +466,6 @@
|
|
469 |
"attn_types": null,
|
470 |
"init_mode": "gpt2",
|
471 |
"normalization_type": null,
|
472 |
-
"device": "cuda",
|
473 |
"n_devices": 1,
|
474 |
"attention_dir": "causal",
|
475 |
"attn_only": false,
|
@@ -482,7 +478,7 @@
|
|
482 |
"d_vocab_out": 2,
|
483 |
"parallel_attn_mlp": false,
|
484 |
"rotary_dim": null,
|
485 |
-
"n_params":
|
486 |
"use_hook_tokens": false,
|
487 |
"gated_mlp": false,
|
488 |
"default_prepend_bos": true,
|
@@ -494,21 +490,22 @@
|
|
494 |
"trust_remote_code": false,
|
495 |
"rotary_adjacent_pairs": false
|
496 |
},
|
497 |
-
"transformer_cfg_file_url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/20/
|
498 |
"training_args": {
|
499 |
"atol": 0.05,
|
500 |
-
"lr": 0.
|
501 |
"use_single_loss": false,
|
502 |
"iit_weight": 1.0,
|
503 |
"behavior_weight": 1.0,
|
504 |
-
"strict_weight":
|
505 |
"epochs": 2000,
|
506 |
"act_fn": "gelu",
|
507 |
"clip_grad_norm": 0.1,
|
508 |
-
"lr_scheduler": ""
|
|
|
509 |
},
|
510 |
-
"training_args_file_url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/20/
|
511 |
-
"weights_file_url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/20/
|
512 |
"circuit_file_url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/20/edges.pkl"
|
513 |
},
|
514 |
{
|
@@ -528,26 +525,26 @@
|
|
528 |
"url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/21/edges.pkl"
|
529 |
},
|
530 |
{
|
531 |
-
"file_name": "
|
532 |
-
"url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/21/
|
533 |
},
|
534 |
{
|
535 |
-
"file_name": "
|
536 |
-
"url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/21/
|
537 |
},
|
538 |
{
|
539 |
-
"file_name": "
|
540 |
-
"url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/21/
|
541 |
}
|
542 |
],
|
543 |
"transformer_cfg": {
|
544 |
-
"n_layers":
|
545 |
-
"d_model":
|
546 |
"n_ctx": 10,
|
547 |
-
"d_head":
|
548 |
"model_name": "custom",
|
549 |
"n_heads": 4,
|
550 |
-
"d_mlp":
|
551 |
"act_fn": "gelu",
|
552 |
"d_vocab": 5,
|
553 |
"eps": 1e-05,
|
@@ -567,12 +564,11 @@
|
|
567 |
"attn_types": null,
|
568 |
"init_mode": "gpt2",
|
569 |
"normalization_type": null,
|
570 |
-
"device": "cpu",
|
571 |
"n_devices": 1,
|
572 |
"attention_dir": "causal",
|
573 |
"attn_only": false,
|
574 |
"seed": 0,
|
575 |
-
"initializer_range": 0.
|
576 |
"init_weights": true,
|
577 |
"scale_attn_by_inverse_layer_idx": false,
|
578 |
"positional_embedding_type": "standard",
|
@@ -580,7 +576,7 @@
|
|
580 |
"d_vocab_out": 3,
|
581 |
"parallel_attn_mlp": false,
|
582 |
"rotary_dim": null,
|
583 |
-
"n_params":
|
584 |
"use_hook_tokens": false,
|
585 |
"gated_mlp": false,
|
586 |
"default_prepend_bos": true,
|
@@ -592,27 +588,28 @@
|
|
592 |
"trust_remote_code": false,
|
593 |
"rotary_adjacent_pairs": false
|
594 |
},
|
595 |
-
"transformer_cfg_file_url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/21/
|
596 |
"training_args": {
|
597 |
"atol": 0.05,
|
598 |
-
"lr": 0.
|
599 |
"use_single_loss": false,
|
600 |
"iit_weight": 1.0,
|
601 |
"behavior_weight": 1.0,
|
602 |
-
"strict_weight": 0.
|
603 |
-
"epochs":
|
604 |
"act_fn": "gelu",
|
605 |
-
"clip_grad_norm": 1
|
606 |
-
"lr_scheduler": ""
|
|
|
607 |
},
|
608 |
-
"training_args_file_url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/21/
|
609 |
-
"weights_file_url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/21/
|
610 |
"circuit_file_url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/21/edges.pkl"
|
611 |
},
|
612 |
{
|
613 |
-
"case_id": "
|
614 |
-
"url": "https://huggingface.co/cybershiptrooper/InterpBench/tree/main/
|
615 |
-
"task_description": "
|
616 |
"vocab": [
|
617 |
"a",
|
618 |
"b",
|
@@ -623,29 +620,29 @@
|
|
623 |
"files": [
|
624 |
{
|
625 |
"file_name": "edges.pkl",
|
626 |
-
"url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/
|
627 |
},
|
628 |
{
|
629 |
-
"file_name": "
|
630 |
-
"url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/
|
631 |
},
|
632 |
{
|
633 |
-
"file_name": "
|
634 |
-
"url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/
|
635 |
},
|
636 |
{
|
637 |
-
"file_name": "
|
638 |
-
"url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/
|
639 |
}
|
640 |
],
|
641 |
"transformer_cfg": {
|
642 |
"n_layers": 2,
|
643 |
-
"d_model":
|
644 |
"n_ctx": 10,
|
645 |
"d_head": 5,
|
646 |
"model_name": "custom",
|
647 |
"n_heads": 4,
|
648 |
-
"d_mlp":
|
649 |
"act_fn": "gelu",
|
650 |
"d_vocab": 5,
|
651 |
"eps": 1e-05,
|
@@ -665,20 +662,19 @@
|
|
665 |
"attn_types": null,
|
666 |
"init_mode": "gpt2",
|
667 |
"normalization_type": null,
|
668 |
-
"device": "cpu",
|
669 |
"n_devices": 1,
|
670 |
"attention_dir": "causal",
|
671 |
"attn_only": false,
|
672 |
"seed": 0,
|
673 |
-
"initializer_range": 0.
|
674 |
"init_weights": true,
|
675 |
"scale_attn_by_inverse_layer_idx": false,
|
676 |
"positional_embedding_type": "standard",
|
677 |
"final_rms": false,
|
678 |
-
"d_vocab_out":
|
679 |
"parallel_attn_mlp": false,
|
680 |
"rotary_dim": null,
|
681 |
-
"n_params":
|
682 |
"use_hook_tokens": false,
|
683 |
"gated_mlp": false,
|
684 |
"default_prepend_bos": true,
|
@@ -690,7 +686,7 @@
|
|
690 |
"trust_remote_code": false,
|
691 |
"rotary_adjacent_pairs": false
|
692 |
},
|
693 |
-
"transformer_cfg_file_url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/
|
694 |
"training_args": {
|
695 |
"atol": 0.05,
|
696 |
"lr": 0.01,
|
@@ -698,14 +694,118 @@
|
|
698 |
"iit_weight": 1.0,
|
699 |
"behavior_weight": 1.0,
|
700 |
"strict_weight": 0.4,
|
701 |
-
"epochs":
|
702 |
"act_fn": "gelu",
|
703 |
-
"clip_grad_norm": 1
|
704 |
-
"lr_scheduler": ""
|
|
|
705 |
},
|
706 |
-
"training_args_file_url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/
|
707 |
-
"weights_file_url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/
|
708 |
-
"circuit_file_url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
709 |
},
|
710 |
{
|
711 |
"case_id": "3",
|
@@ -725,16 +825,16 @@
|
|
725 |
"url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/3/edges.pkl"
|
726 |
},
|
727 |
{
|
728 |
-
"file_name": "
|
729 |
-
"url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/3/
|
730 |
},
|
731 |
{
|
732 |
-
"file_name": "
|
733 |
-
"url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/3/
|
734 |
},
|
735 |
{
|
736 |
-
"file_name": "
|
737 |
-
"url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/3/
|
738 |
}
|
739 |
],
|
740 |
"transformer_cfg": {
|
@@ -764,7 +864,6 @@
|
|
764 |
"attn_types": null,
|
765 |
"init_mode": "gpt2",
|
766 |
"normalization_type": null,
|
767 |
-
"device": "cpu",
|
768 |
"n_devices": 1,
|
769 |
"attention_dir": "causal",
|
770 |
"attn_only": false,
|
@@ -789,7 +888,7 @@
|
|
789 |
"trust_remote_code": false,
|
790 |
"rotary_adjacent_pairs": false
|
791 |
},
|
792 |
-
"transformer_cfg_file_url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/3/
|
793 |
"training_args": {
|
794 |
"atol": 0.05,
|
795 |
"lr": 0.001,
|
@@ -802,8 +901,8 @@
|
|
802 |
"clip_grad_norm": 0.1,
|
803 |
"lr_scheduler": ""
|
804 |
},
|
805 |
-
"training_args_file_url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/3/
|
806 |
-
"weights_file_url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/3/
|
807 |
"circuit_file_url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/3/edges.pkl"
|
808 |
},
|
809 |
{
|
@@ -828,16 +927,16 @@
|
|
828 |
"url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/33/edges.pkl"
|
829 |
},
|
830 |
{
|
831 |
-
"file_name": "
|
832 |
-
"url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/33/
|
833 |
},
|
834 |
{
|
835 |
-
"file_name": "
|
836 |
-
"url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/33/
|
837 |
},
|
838 |
{
|
839 |
-
"file_name": "
|
840 |
-
"url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/33/
|
841 |
}
|
842 |
],
|
843 |
"transformer_cfg": {
|
@@ -867,7 +966,6 @@
|
|
867 |
"attn_types": null,
|
868 |
"init_mode": "gpt2",
|
869 |
"normalization_type": null,
|
870 |
-
"device": "cpu",
|
871 |
"n_devices": 1,
|
872 |
"attention_dir": "causal",
|
873 |
"attn_only": false,
|
@@ -892,7 +990,7 @@
|
|
892 |
"trust_remote_code": false,
|
893 |
"rotary_adjacent_pairs": false
|
894 |
},
|
895 |
-
"transformer_cfg_file_url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/33/
|
896 |
"training_args": {
|
897 |
"atol": 0.05,
|
898 |
"lr": 0.001,
|
@@ -905,8 +1003,8 @@
|
|
905 |
"clip_grad_norm": 0.1,
|
906 |
"lr_scheduler": ""
|
907 |
},
|
908 |
-
"training_args_file_url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/33/
|
909 |
-
"weights_file_url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/33/
|
910 |
"circuit_file_url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/33/edges.pkl"
|
911 |
},
|
912 |
{
|
@@ -931,26 +1029,26 @@
|
|
931 |
"url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/34/edges.pkl"
|
932 |
},
|
933 |
{
|
934 |
-
"file_name": "
|
935 |
-
"url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/34/
|
936 |
},
|
937 |
{
|
938 |
-
"file_name": "
|
939 |
-
"url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/34/
|
940 |
},
|
941 |
{
|
942 |
-
"file_name": "
|
943 |
-
"url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/34/
|
944 |
}
|
945 |
],
|
946 |
"transformer_cfg": {
|
947 |
"n_layers": 2,
|
948 |
-
"d_model":
|
949 |
"n_ctx": 10,
|
950 |
-
"d_head":
|
951 |
"model_name": "custom",
|
952 |
"n_heads": 4,
|
953 |
-
"d_mlp":
|
954 |
"act_fn": "gelu",
|
955 |
"d_vocab": 10,
|
956 |
"eps": 1e-05,
|
@@ -970,7 +1068,6 @@
|
|
970 |
"attn_types": null,
|
971 |
"init_mode": "gpt2",
|
972 |
"normalization_type": null,
|
973 |
-
"device": "cpu",
|
974 |
"n_devices": 1,
|
975 |
"attention_dir": "causal",
|
976 |
"attn_only": false,
|
@@ -983,7 +1080,7 @@
|
|
983 |
"d_vocab_out": 5,
|
984 |
"parallel_attn_mlp": false,
|
985 |
"rotary_dim": null,
|
986 |
-
"n_params":
|
987 |
"use_hook_tokens": false,
|
988 |
"gated_mlp": false,
|
989 |
"default_prepend_bos": true,
|
@@ -995,21 +1092,22 @@
|
|
995 |
"trust_remote_code": false,
|
996 |
"rotary_adjacent_pairs": false
|
997 |
},
|
998 |
-
"transformer_cfg_file_url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/34/
|
999 |
"training_args": {
|
1000 |
"atol": 0.05,
|
1001 |
-
"lr": 0.
|
1002 |
"use_single_loss": false,
|
1003 |
"iit_weight": 1.0,
|
1004 |
"behavior_weight": 1.0,
|
1005 |
-
"strict_weight": 0
|
1006 |
"epochs": 2000,
|
1007 |
"act_fn": "gelu",
|
1008 |
"clip_grad_norm": 0.1,
|
1009 |
-
"lr_scheduler": ""
|
|
|
1010 |
},
|
1011 |
-
"training_args_file_url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/34/
|
1012 |
-
"weights_file_url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/34/
|
1013 |
"circuit_file_url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/34/edges.pkl"
|
1014 |
},
|
1015 |
{
|
@@ -1034,26 +1132,26 @@
|
|
1034 |
"url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/35/edges.pkl"
|
1035 |
},
|
1036 |
{
|
1037 |
-
"file_name": "
|
1038 |
-
"url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/35/
|
1039 |
},
|
1040 |
{
|
1041 |
-
"file_name": "
|
1042 |
-
"url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/35/
|
1043 |
},
|
1044 |
{
|
1045 |
-
"file_name": "
|
1046 |
-
"url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/35/
|
1047 |
}
|
1048 |
],
|
1049 |
"transformer_cfg": {
|
1050 |
"n_layers": 2,
|
1051 |
-
"d_model":
|
1052 |
"n_ctx": 10,
|
1053 |
-
"d_head":
|
1054 |
"model_name": "custom",
|
1055 |
"n_heads": 4,
|
1056 |
-
"d_mlp":
|
1057 |
"act_fn": "gelu",
|
1058 |
"d_vocab": 10,
|
1059 |
"eps": 1e-05,
|
@@ -1073,7 +1171,6 @@
|
|
1073 |
"attn_types": null,
|
1074 |
"init_mode": "gpt2",
|
1075 |
"normalization_type": null,
|
1076 |
-
"device": "cpu",
|
1077 |
"n_devices": 1,
|
1078 |
"attention_dir": "causal",
|
1079 |
"attn_only": false,
|
@@ -1086,7 +1183,7 @@
|
|
1086 |
"d_vocab_out": 8,
|
1087 |
"parallel_attn_mlp": false,
|
1088 |
"rotary_dim": null,
|
1089 |
-
"n_params":
|
1090 |
"use_hook_tokens": false,
|
1091 |
"gated_mlp": false,
|
1092 |
"default_prepend_bos": true,
|
@@ -1098,21 +1195,22 @@
|
|
1098 |
"trust_remote_code": false,
|
1099 |
"rotary_adjacent_pairs": false
|
1100 |
},
|
1101 |
-
"transformer_cfg_file_url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/35/
|
1102 |
"training_args": {
|
1103 |
"atol": 0.05,
|
1104 |
-
"lr": 0.
|
1105 |
"use_single_loss": false,
|
1106 |
"iit_weight": 1.0,
|
1107 |
"behavior_weight": 1.0,
|
1108 |
-
"strict_weight": 0
|
1109 |
"epochs": 2000,
|
1110 |
"act_fn": "gelu",
|
1111 |
"clip_grad_norm": 0.1,
|
1112 |
-
"lr_scheduler": ""
|
|
|
1113 |
},
|
1114 |
-
"training_args_file_url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/35/
|
1115 |
-
"weights_file_url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/35/
|
1116 |
"circuit_file_url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/35/edges.pkl"
|
1117 |
},
|
1118 |
{
|
@@ -1132,26 +1230,26 @@
|
|
1132 |
"url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/36/edges.pkl"
|
1133 |
},
|
1134 |
{
|
1135 |
-
"file_name": "
|
1136 |
-
"url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/36/
|
1137 |
},
|
1138 |
{
|
1139 |
-
"file_name": "
|
1140 |
-
"url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/36/
|
1141 |
},
|
1142 |
{
|
1143 |
-
"file_name": "
|
1144 |
-
"url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/36/
|
1145 |
}
|
1146 |
],
|
1147 |
"transformer_cfg": {
|
1148 |
"n_layers": 2,
|
1149 |
-
"d_model":
|
1150 |
"n_ctx": 10,
|
1151 |
"d_head": 1,
|
1152 |
"model_name": "custom",
|
1153 |
"n_heads": 4,
|
1154 |
-
"d_mlp":
|
1155 |
"act_fn": "gelu",
|
1156 |
"d_vocab": 5,
|
1157 |
"eps": 1e-05,
|
@@ -1171,7 +1269,6 @@
|
|
1171 |
"attn_types": null,
|
1172 |
"init_mode": "gpt2",
|
1173 |
"normalization_type": null,
|
1174 |
-
"device": "cuda",
|
1175 |
"n_devices": 1,
|
1176 |
"attention_dir": "causal",
|
1177 |
"attn_only": false,
|
@@ -1184,7 +1281,7 @@
|
|
1184 |
"d_vocab_out": 3,
|
1185 |
"parallel_attn_mlp": false,
|
1186 |
"rotary_dim": null,
|
1187 |
-
"n_params":
|
1188 |
"use_hook_tokens": false,
|
1189 |
"gated_mlp": false,
|
1190 |
"default_prepend_bos": true,
|
@@ -1196,21 +1293,22 @@
|
|
1196 |
"trust_remote_code": false,
|
1197 |
"rotary_adjacent_pairs": false
|
1198 |
},
|
1199 |
-
"transformer_cfg_file_url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/36/
|
1200 |
"training_args": {
|
1201 |
"atol": 0.05,
|
1202 |
-
"lr": 0.
|
1203 |
"use_single_loss": false,
|
1204 |
"iit_weight": 1.0,
|
1205 |
"behavior_weight": 1.0,
|
1206 |
-
"strict_weight":
|
1207 |
"epochs": 2000,
|
1208 |
"act_fn": "gelu",
|
1209 |
"clip_grad_norm": 0.1,
|
1210 |
-
"lr_scheduler": ""
|
|
|
1211 |
},
|
1212 |
-
"training_args_file_url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/36/
|
1213 |
-
"weights_file_url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/36/
|
1214 |
"circuit_file_url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/36/edges.pkl"
|
1215 |
},
|
1216 |
{
|
@@ -1235,26 +1333,26 @@
|
|
1235 |
"url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/37/edges.pkl"
|
1236 |
},
|
1237 |
{
|
1238 |
-
"file_name": "
|
1239 |
-
"url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/37/
|
1240 |
},
|
1241 |
{
|
1242 |
-
"file_name": "
|
1243 |
-
"url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/37/
|
1244 |
},
|
1245 |
{
|
1246 |
-
"file_name": "
|
1247 |
-
"url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/37/
|
1248 |
}
|
1249 |
],
|
1250 |
"transformer_cfg": {
|
1251 |
"n_layers": 2,
|
1252 |
-
"d_model":
|
1253 |
"n_ctx": 10,
|
1254 |
-
"d_head":
|
1255 |
"model_name": "custom",
|
1256 |
"n_heads": 4,
|
1257 |
-
"d_mlp":
|
1258 |
"act_fn": "gelu",
|
1259 |
"d_vocab": 10,
|
1260 |
"eps": 1e-05,
|
@@ -1274,7 +1372,6 @@
|
|
1274 |
"attn_types": null,
|
1275 |
"init_mode": "gpt2",
|
1276 |
"normalization_type": null,
|
1277 |
-
"device": "cpu",
|
1278 |
"n_devices": 1,
|
1279 |
"attention_dir": "causal",
|
1280 |
"attn_only": false,
|
@@ -1287,7 +1384,7 @@
|
|
1287 |
"d_vocab_out": 8,
|
1288 |
"parallel_attn_mlp": false,
|
1289 |
"rotary_dim": null,
|
1290 |
-
"n_params":
|
1291 |
"use_hook_tokens": false,
|
1292 |
"gated_mlp": false,
|
1293 |
"default_prepend_bos": true,
|
@@ -1299,28 +1396,31 @@
|
|
1299 |
"trust_remote_code": false,
|
1300 |
"rotary_adjacent_pairs": false
|
1301 |
},
|
1302 |
-
"transformer_cfg_file_url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/37/
|
1303 |
"training_args": {
|
1304 |
"atol": 0.05,
|
1305 |
-
"lr": 0.
|
1306 |
"use_single_loss": false,
|
1307 |
"iit_weight": 1.0,
|
1308 |
"behavior_weight": 1.0,
|
1309 |
-
"strict_weight": 0
|
1310 |
"epochs": 2000,
|
1311 |
"act_fn": "gelu",
|
1312 |
"clip_grad_norm": 0.1,
|
1313 |
-
"lr_scheduler": ""
|
|
|
1314 |
},
|
1315 |
-
"training_args_file_url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/37/
|
1316 |
-
"weights_file_url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/37/
|
1317 |
"circuit_file_url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/37/edges.pkl"
|
1318 |
},
|
1319 |
{
|
1320 |
-
"case_id": "
|
1321 |
-
"url": "https://huggingface.co/cybershiptrooper/InterpBench/tree/main/
|
1322 |
-
"task_description": "
|
1323 |
"vocab": [
|
|
|
|
|
1324 |
"a",
|
1325 |
"b",
|
1326 |
"c"
|
@@ -1330,19 +1430,19 @@
|
|
1330 |
"files": [
|
1331 |
{
|
1332 |
"file_name": "edges.pkl",
|
1333 |
-
"url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/
|
1334 |
},
|
1335 |
{
|
1336 |
-
"file_name": "
|
1337 |
-
"url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/
|
1338 |
},
|
1339 |
{
|
1340 |
-
"file_name": "
|
1341 |
-
"url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/
|
1342 |
},
|
1343 |
{
|
1344 |
-
"file_name": "
|
1345 |
-
"url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/
|
1346 |
}
|
1347 |
],
|
1348 |
"transformer_cfg": {
|
@@ -1354,7 +1454,7 @@
|
|
1354 |
"n_heads": 4,
|
1355 |
"d_mlp": 80,
|
1356 |
"act_fn": "gelu",
|
1357 |
-
"d_vocab":
|
1358 |
"eps": 1e-05,
|
1359 |
"use_attn_result": true,
|
1360 |
"use_attn_scale": true,
|
@@ -1372,17 +1472,16 @@
|
|
1372 |
"attn_types": null,
|
1373 |
"init_mode": "gpt2",
|
1374 |
"normalization_type": null,
|
1375 |
-
"device": "cpu",
|
1376 |
"n_devices": 1,
|
1377 |
"attention_dir": "causal",
|
1378 |
"attn_only": false,
|
1379 |
"seed": 0,
|
1380 |
-
"initializer_range": 0.
|
1381 |
"init_weights": true,
|
1382 |
"scale_attn_by_inverse_layer_idx": false,
|
1383 |
"positional_embedding_type": "standard",
|
1384 |
"final_rms": false,
|
1385 |
-
"d_vocab_out":
|
1386 |
"parallel_attn_mlp": false,
|
1387 |
"rotary_dim": null,
|
1388 |
"n_params": 9600,
|
@@ -1397,7 +1496,7 @@
|
|
1397 |
"trust_remote_code": false,
|
1398 |
"rotary_adjacent_pairs": false
|
1399 |
},
|
1400 |
-
"transformer_cfg_file_url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/
|
1401 |
"training_args": {
|
1402 |
"atol": 0.05,
|
1403 |
"lr": 0.001,
|
@@ -1410,39 +1509,42 @@
|
|
1410 |
"clip_grad_norm": 0.1,
|
1411 |
"lr_scheduler": ""
|
1412 |
},
|
1413 |
-
"training_args_file_url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/
|
1414 |
-
"weights_file_url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/
|
1415 |
-
"circuit_file_url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/
|
1416 |
},
|
1417 |
{
|
1418 |
-
"case_id": "
|
1419 |
-
"url": "https://huggingface.co/cybershiptrooper/InterpBench/tree/main/
|
1420 |
-
"task_description": "
|
1421 |
"vocab": [
|
1422 |
-
"
|
1423 |
-
"
|
1424 |
-
"
|
|
|
1425 |
"b",
|
1426 |
-
"
|
|
|
|
|
1427 |
],
|
1428 |
"max_seq_len": 10,
|
1429 |
"min_seq_len": 4,
|
1430 |
"files": [
|
1431 |
{
|
1432 |
"file_name": "edges.pkl",
|
1433 |
-
"url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/
|
1434 |
},
|
1435 |
{
|
1436 |
-
"file_name": "
|
1437 |
-
"url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/
|
1438 |
},
|
1439 |
{
|
1440 |
-
"file_name": "
|
1441 |
-
"url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/
|
1442 |
},
|
1443 |
{
|
1444 |
-
"file_name": "
|
1445 |
-
"url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/
|
1446 |
}
|
1447 |
],
|
1448 |
"transformer_cfg": {
|
@@ -1454,7 +1556,7 @@
|
|
1454 |
"n_heads": 4,
|
1455 |
"d_mlp": 80,
|
1456 |
"act_fn": "gelu",
|
1457 |
-
"d_vocab":
|
1458 |
"eps": 1e-05,
|
1459 |
"use_attn_result": true,
|
1460 |
"use_attn_scale": true,
|
@@ -1472,17 +1574,16 @@
|
|
1472 |
"attn_types": null,
|
1473 |
"init_mode": "gpt2",
|
1474 |
"normalization_type": null,
|
1475 |
-
"device": "cpu",
|
1476 |
"n_devices": 1,
|
1477 |
"attention_dir": "causal",
|
1478 |
"attn_only": false,
|
1479 |
"seed": 0,
|
1480 |
-
"initializer_range": 0.
|
1481 |
"init_weights": true,
|
1482 |
"scale_attn_by_inverse_layer_idx": false,
|
1483 |
"positional_embedding_type": "standard",
|
1484 |
"final_rms": false,
|
1485 |
-
"d_vocab_out":
|
1486 |
"parallel_attn_mlp": false,
|
1487 |
"rotary_dim": null,
|
1488 |
"n_params": 9600,
|
@@ -1497,143 +1598,123 @@
|
|
1497 |
"trust_remote_code": false,
|
1498 |
"rotary_adjacent_pairs": false
|
1499 |
},
|
1500 |
-
"transformer_cfg_file_url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/
|
1501 |
"training_args": {
|
1502 |
"atol": 0.05,
|
1503 |
-
"lr": 0.
|
1504 |
"use_single_loss": false,
|
1505 |
"iit_weight": 1.0,
|
1506 |
"behavior_weight": 1.0,
|
1507 |
"strict_weight": 0.4,
|
1508 |
-
"epochs":
|
1509 |
"act_fn": "gelu",
|
1510 |
-
"clip_grad_norm": 0
|
1511 |
"lr_scheduler": ""
|
1512 |
},
|
1513 |
-
"training_args_file_url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/
|
1514 |
-
"weights_file_url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/
|
1515 |
-
"circuit_file_url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/
|
1516 |
},
|
1517 |
{
|
1518 |
-
"case_id": "
|
1519 |
-
"url": "https://huggingface.co/cybershiptrooper/InterpBench/tree/main/
|
1520 |
-
"task_description": "
|
1521 |
-
"
|
1522 |
-
|
1523 |
-
"LB",
|
1524 |
-
"TPSI",
|
1525 |
-
"V",
|
1526 |
-
"b",
|
1527 |
-
"no",
|
1528 |
-
"oCLrZaW",
|
1529 |
-
"poiVg"
|
1530 |
-
],
|
1531 |
-
"max_seq_len": 10,
|
1532 |
-
"min_seq_len": 4,
|
1533 |
"files": [
|
1534 |
{
|
1535 |
"file_name": "edges.pkl",
|
1536 |
-
"url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/
|
1537 |
},
|
1538 |
{
|
1539 |
-
"file_name": "
|
1540 |
-
"url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/
|
1541 |
},
|
1542 |
{
|
1543 |
-
"file_name": "
|
1544 |
-
"url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/
|
1545 |
},
|
1546 |
{
|
1547 |
-
"file_name": "
|
1548 |
-
"url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/
|
1549 |
}
|
1550 |
],
|
1551 |
"transformer_cfg": {
|
1552 |
-
"n_layers":
|
1553 |
-
"d_model":
|
1554 |
-
"n_ctx":
|
1555 |
-
"d_head":
|
1556 |
-
"model_name": "
|
1557 |
"n_heads": 4,
|
1558 |
-
"d_mlp":
|
1559 |
-
"act_fn": "
|
1560 |
-
"d_vocab":
|
1561 |
"eps": 1e-05,
|
1562 |
-
"use_attn_result":
|
1563 |
"use_attn_scale": true,
|
1564 |
-
"use_split_qkv_input":
|
1565 |
-
"use_hook_mlp_in":
|
1566 |
"use_attn_in": false,
|
1567 |
"use_local_attn": false,
|
1568 |
-
"original_architecture":
|
1569 |
"from_checkpoint": false,
|
1570 |
"checkpoint_index": null,
|
1571 |
"checkpoint_label_type": null,
|
1572 |
"checkpoint_value": null,
|
1573 |
-
"tokenizer_name":
|
1574 |
"window_size": null,
|
1575 |
"attn_types": null,
|
1576 |
"init_mode": "gpt2",
|
1577 |
-
"normalization_type":
|
1578 |
-
"device": "cpu",
|
1579 |
"n_devices": 1,
|
1580 |
"attention_dir": "causal",
|
1581 |
"attn_only": false,
|
1582 |
-
"seed":
|
1583 |
-
"initializer_range": 0.
|
1584 |
"init_weights": true,
|
1585 |
"scale_attn_by_inverse_layer_idx": false,
|
1586 |
"positional_embedding_type": "standard",
|
1587 |
"final_rms": false,
|
1588 |
-
"d_vocab_out":
|
1589 |
"parallel_attn_mlp": false,
|
1590 |
"rotary_dim": null,
|
1591 |
-
"n_params":
|
1592 |
"use_hook_tokens": false,
|
1593 |
"gated_mlp": false,
|
1594 |
"default_prepend_bos": true,
|
1595 |
"dtype": "torch.float32",
|
1596 |
-
"tokenizer_prepends_bos":
|
1597 |
"n_key_value_heads": null,
|
1598 |
"post_embedding_ln": false,
|
1599 |
"rotary_base": 10000,
|
1600 |
"trust_remote_code": false,
|
1601 |
"rotary_adjacent_pairs": false
|
1602 |
},
|
1603 |
-
"transformer_cfg_file_url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/
|
1604 |
"training_args": {
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1605 |
"atol": 0.05,
|
1606 |
-
"lr": 0.01,
|
1607 |
"use_single_loss": false,
|
1608 |
"iit_weight": 1.0,
|
1609 |
"behavior_weight": 1.0,
|
1610 |
-
"strict_weight": 0.4
|
1611 |
-
"epochs": 500,
|
1612 |
-
"act_fn": "gelu",
|
1613 |
-
"clip_grad_norm": 1.0,
|
1614 |
-
"lr_scheduler": ""
|
1615 |
},
|
1616 |
-
"training_args_file_url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/
|
1617 |
-
"weights_file_url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/
|
1618 |
-
"circuit_file_url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/
|
1619 |
-
},
|
1620 |
-
{
|
1621 |
-
"case_id": "ioi",
|
1622 |
-
"url": "https://huggingface.co/cybershiptrooper/InterpBench/tree/main/ioi",
|
1623 |
-
"task_description": "Indirect object identification",
|
1624 |
-
"max_seq_len": 16,
|
1625 |
-
"min_seq_len": 16,
|
1626 |
-
"files": [
|
1627 |
-
{
|
1628 |
-
"file_name": "corr_100_100_40.json",
|
1629 |
-
"url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/ioi/corr_100_100_40.json"
|
1630 |
-
},
|
1631 |
-
{
|
1632 |
-
"file_name": "ll_model_100_100_40.pth",
|
1633 |
-
"url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/ioi/ll_model_100_100_40.pth"
|
1634 |
-
}
|
1635 |
-
],
|
1636 |
-
"weights_file_url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/ioi/ll_model_100_100_40.pth"
|
1637 |
},
|
1638 |
{
|
1639 |
"case_id": "ioi_next_token",
|
@@ -1643,19 +1724,98 @@
|
|
1643 |
"min_seq_len": 16,
|
1644 |
"files": [
|
1645 |
{
|
1646 |
-
"file_name": "
|
1647 |
-
"url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/ioi_next_token/
|
|
|
|
|
|
|
|
|
1648 |
},
|
1649 |
{
|
1650 |
-
"file_name": "
|
1651 |
-
"url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/ioi_next_token/
|
1652 |
},
|
1653 |
{
|
1654 |
-
"file_name": "
|
1655 |
-
"url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/ioi_next_token/
|
1656 |
}
|
1657 |
],
|
1658 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1659 |
}
|
1660 |
]
|
1661 |
}
|
|
|
27 |
"url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/11/edges.pkl"
|
28 |
},
|
29 |
{
|
30 |
+
"file_name": "ll_model.pth",
|
31 |
+
"url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/11/ll_model.pth"
|
32 |
},
|
33 |
{
|
34 |
+
"file_name": "ll_model_cfg.pkl",
|
35 |
+
"url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/11/ll_model_cfg.pkl"
|
36 |
},
|
37 |
{
|
38 |
+
"file_name": "meta.json",
|
39 |
+
"url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/11/meta.json"
|
40 |
}
|
41 |
],
|
42 |
"transformer_cfg": {
|
|
|
66 |
"attn_types": null,
|
67 |
"init_mode": "gpt2",
|
68 |
"normalization_type": null,
|
|
|
69 |
"n_devices": 1,
|
70 |
"attention_dir": "causal",
|
71 |
"attn_only": false,
|
|
|
90 |
"trust_remote_code": false,
|
91 |
"rotary_adjacent_pairs": false
|
92 |
},
|
93 |
+
"transformer_cfg_file_url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/11/ll_model_cfg.pkl",
|
94 |
"training_args": {
|
95 |
"atol": 0.05,
|
96 |
"lr": 0.01,
|
|
|
103 |
"clip_grad_norm": 1.0,
|
104 |
"lr_scheduler": ""
|
105 |
},
|
106 |
+
"training_args_file_url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/11/meta.json",
|
107 |
+
"weights_file_url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/11/ll_model.pth",
|
108 |
"circuit_file_url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/11/edges.pkl"
|
109 |
},
|
110 |
{
|
|
|
124 |
"url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/13/edges.pkl"
|
125 |
},
|
126 |
{
|
127 |
+
"file_name": "ll_model.pth",
|
128 |
+
"url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/13/ll_model.pth"
|
129 |
},
|
130 |
{
|
131 |
+
"file_name": "ll_model_cfg.pkl",
|
132 |
+
"url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/13/ll_model_cfg.pkl"
|
133 |
},
|
134 |
{
|
135 |
+
"file_name": "meta.json",
|
136 |
+
"url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/13/meta.json"
|
137 |
}
|
138 |
],
|
139 |
"transformer_cfg": {
|
|
|
163 |
"attn_types": null,
|
164 |
"init_mode": "gpt2",
|
165 |
"normalization_type": null,
|
|
|
166 |
"n_devices": 1,
|
167 |
"attention_dir": "bidirectional",
|
168 |
"attn_only": false,
|
|
|
187 |
"trust_remote_code": false,
|
188 |
"rotary_adjacent_pairs": false
|
189 |
},
|
190 |
+
"transformer_cfg_file_url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/13/ll_model_cfg.pkl",
|
191 |
"training_args": {
|
192 |
"atol": 0.05,
|
193 |
"lr": 0.01,
|
|
|
200 |
"clip_grad_norm": 1.0,
|
201 |
"lr_scheduler": ""
|
202 |
},
|
203 |
+
"training_args_file_url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/13/meta.json",
|
204 |
+
"weights_file_url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/13/ll_model.pth",
|
205 |
"circuit_file_url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/13/edges.pkl"
|
206 |
},
|
207 |
{
|
|
|
223 |
"url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/18/edges.pkl"
|
224 |
},
|
225 |
{
|
226 |
+
"file_name": "ll_model.pth",
|
227 |
+
"url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/18/ll_model.pth"
|
228 |
},
|
229 |
{
|
230 |
+
"file_name": "ll_model_cfg.pkl",
|
231 |
+
"url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/18/ll_model_cfg.pkl"
|
232 |
},
|
233 |
{
|
234 |
+
"file_name": "meta.json",
|
235 |
+
"url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/18/meta.json"
|
236 |
}
|
237 |
],
|
238 |
"transformer_cfg": {
|
239 |
"n_layers": 2,
|
240 |
+
"d_model": 26,
|
241 |
"n_ctx": 10,
|
242 |
+
"d_head": 6,
|
243 |
"model_name": "custom",
|
244 |
"n_heads": 4,
|
245 |
+
"d_mlp": 104,
|
246 |
"act_fn": "gelu",
|
247 |
"d_vocab": 7,
|
248 |
"eps": 1e-05,
|
|
|
262 |
"attn_types": null,
|
263 |
"init_mode": "gpt2",
|
264 |
"normalization_type": null,
|
|
|
265 |
"n_devices": 1,
|
266 |
"attention_dir": "bidirectional",
|
267 |
"attn_only": false,
|
|
|
274 |
"d_vocab_out": 3,
|
275 |
"parallel_attn_mlp": false,
|
276 |
"rotary_dim": null,
|
277 |
+
"n_params": 15808,
|
278 |
"use_hook_tokens": false,
|
279 |
"gated_mlp": false,
|
280 |
"default_prepend_bos": true,
|
|
|
286 |
"trust_remote_code": false,
|
287 |
"rotary_adjacent_pairs": false
|
288 |
},
|
289 |
+
"transformer_cfg_file_url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/18/ll_model_cfg.pkl",
|
290 |
"training_args": {
|
291 |
"atol": 0.05,
|
292 |
"lr": 0.001,
|
293 |
"use_single_loss": false,
|
294 |
"iit_weight": 1.0,
|
295 |
"behavior_weight": 1.0,
|
296 |
+
"strict_weight": 1.0,
|
297 |
"epochs": 2000,
|
298 |
"act_fn": "gelu",
|
299 |
"clip_grad_norm": 0.1,
|
300 |
+
"lr_scheduler": "",
|
301 |
+
"model_pair": "strict"
|
302 |
},
|
303 |
+
"training_args_file_url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/18/meta.json",
|
304 |
+
"weights_file_url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/18/ll_model.pth",
|
305 |
"circuit_file_url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/18/edges.pkl"
|
306 |
},
|
307 |
{
|
|
|
321 |
"url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/19/edges.pkl"
|
322 |
},
|
323 |
{
|
324 |
+
"file_name": "ll_model.pth",
|
325 |
+
"url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/19/ll_model.pth"
|
326 |
},
|
327 |
{
|
328 |
+
"file_name": "ll_model_cfg.pkl",
|
329 |
+
"url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/19/ll_model_cfg.pkl"
|
330 |
},
|
331 |
{
|
332 |
+
"file_name": "meta.json",
|
333 |
+
"url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/19/meta.json"
|
334 |
}
|
335 |
],
|
336 |
"transformer_cfg": {
|
|
|
360 |
"attn_types": null,
|
361 |
"init_mode": "gpt2",
|
362 |
"normalization_type": null,
|
|
|
363 |
"n_devices": 1,
|
364 |
"attention_dir": "causal",
|
365 |
"attn_only": false,
|
|
|
384 |
"trust_remote_code": false,
|
385 |
"rotary_adjacent_pairs": false
|
386 |
},
|
387 |
+
"transformer_cfg_file_url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/19/ll_model_cfg.pkl",
|
388 |
"training_args": {
|
389 |
"atol": 0.05,
|
390 |
"lr": 0.001,
|
|
|
397 |
"clip_grad_norm": 0.1,
|
398 |
"lr_scheduler": ""
|
399 |
},
|
400 |
+
"training_args_file_url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/19/meta.json",
|
401 |
+
"weights_file_url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/19/ll_model.pth",
|
402 |
"circuit_file_url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/19/edges.pkl"
|
403 |
},
|
404 |
{
|
|
|
427 |
"url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/20/edges.pkl"
|
428 |
},
|
429 |
{
|
430 |
+
"file_name": "ll_model.pth",
|
431 |
+
"url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/20/ll_model.pth"
|
432 |
},
|
433 |
{
|
434 |
+
"file_name": "ll_model_cfg.pkl",
|
435 |
+
"url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/20/ll_model_cfg.pkl"
|
436 |
},
|
437 |
{
|
438 |
+
"file_name": "meta.json",
|
439 |
+
"url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/20/meta.json"
|
440 |
}
|
441 |
],
|
442 |
"transformer_cfg": {
|
443 |
"n_layers": 2,
|
444 |
+
"d_model": 13,
|
445 |
"n_ctx": 10,
|
446 |
+
"d_head": 3,
|
447 |
"model_name": "custom",
|
448 |
"n_heads": 4,
|
449 |
+
"d_mlp": 52,
|
450 |
"act_fn": "gelu",
|
451 |
"d_vocab": 14,
|
452 |
"eps": 1e-05,
|
|
|
466 |
"attn_types": null,
|
467 |
"init_mode": "gpt2",
|
468 |
"normalization_type": null,
|
|
|
469 |
"n_devices": 1,
|
470 |
"attention_dir": "causal",
|
471 |
"attn_only": false,
|
|
|
478 |
"d_vocab_out": 2,
|
479 |
"parallel_attn_mlp": false,
|
480 |
"rotary_dim": null,
|
481 |
+
"n_params": 3952,
|
482 |
"use_hook_tokens": false,
|
483 |
"gated_mlp": false,
|
484 |
"default_prepend_bos": true,
|
|
|
490 |
"trust_remote_code": false,
|
491 |
"rotary_adjacent_pairs": false
|
492 |
},
|
493 |
+
"transformer_cfg_file_url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/20/ll_model_cfg.pkl",
|
494 |
"training_args": {
|
495 |
"atol": 0.05,
|
496 |
+
"lr": 0.01,
|
497 |
"use_single_loss": false,
|
498 |
"iit_weight": 1.0,
|
499 |
"behavior_weight": 1.0,
|
500 |
+
"strict_weight": 0.4,
|
501 |
"epochs": 2000,
|
502 |
"act_fn": "gelu",
|
503 |
"clip_grad_norm": 0.1,
|
504 |
+
"lr_scheduler": "",
|
505 |
+
"model_pair": "strict"
|
506 |
},
|
507 |
+
"training_args_file_url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/20/meta.json",
|
508 |
+
"weights_file_url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/20/ll_model.pth",
|
509 |
"circuit_file_url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/20/edges.pkl"
|
510 |
},
|
511 |
{
|
|
|
525 |
"url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/21/edges.pkl"
|
526 |
},
|
527 |
{
|
528 |
+
"file_name": "ll_model.pth",
|
529 |
+
"url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/21/ll_model.pth"
|
530 |
},
|
531 |
{
|
532 |
+
"file_name": "ll_model_cfg.pkl",
|
533 |
+
"url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/21/ll_model_cfg.pkl"
|
534 |
},
|
535 |
{
|
536 |
+
"file_name": "meta.json",
|
537 |
+
"url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/21/meta.json"
|
538 |
}
|
539 |
],
|
540 |
"transformer_cfg": {
|
541 |
+
"n_layers": 4,
|
542 |
+
"d_model": 50,
|
543 |
"n_ctx": 10,
|
544 |
+
"d_head": 12,
|
545 |
"model_name": "custom",
|
546 |
"n_heads": 4,
|
547 |
+
"d_mlp": 200,
|
548 |
"act_fn": "gelu",
|
549 |
"d_vocab": 5,
|
550 |
"eps": 1e-05,
|
|
|
564 |
"attn_types": null,
|
565 |
"init_mode": "gpt2",
|
566 |
"normalization_type": null,
|
|
|
567 |
"n_devices": 1,
|
568 |
"attention_dir": "causal",
|
569 |
"attn_only": false,
|
570 |
"seed": 0,
|
571 |
+
"initializer_range": 0.09847319278346618,
|
572 |
"init_weights": true,
|
573 |
"scale_attn_by_inverse_layer_idx": false,
|
574 |
"positional_embedding_type": "standard",
|
|
|
576 |
"d_vocab_out": 3,
|
577 |
"parallel_attn_mlp": false,
|
578 |
"rotary_dim": null,
|
579 |
+
"n_params": 118400,
|
580 |
"use_hook_tokens": false,
|
581 |
"gated_mlp": false,
|
582 |
"default_prepend_bos": true,
|
|
|
588 |
"trust_remote_code": false,
|
589 |
"rotary_adjacent_pairs": false
|
590 |
},
|
591 |
+
"transformer_cfg_file_url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/21/ll_model_cfg.pkl",
|
592 |
"training_args": {
|
593 |
"atol": 0.05,
|
594 |
+
"lr": 0.0005,
|
595 |
"use_single_loss": false,
|
596 |
"iit_weight": 1.0,
|
597 |
"behavior_weight": 1.0,
|
598 |
+
"strict_weight": 0.5,
|
599 |
+
"epochs": 2000,
|
600 |
"act_fn": "gelu",
|
601 |
+
"clip_grad_norm": 0.1,
|
602 |
+
"lr_scheduler": "",
|
603 |
+
"model_pair": "strict"
|
604 |
},
|
605 |
+
"training_args_file_url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/21/meta.json",
|
606 |
+
"weights_file_url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/21/ll_model.pth",
|
607 |
"circuit_file_url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/21/edges.pkl"
|
608 |
},
|
609 |
{
|
610 |
+
"case_id": "26",
|
611 |
+
"url": "https://huggingface.co/cybershiptrooper/InterpBench/tree/main/26",
|
612 |
+
"task_description": "Creates a cascading effect by repeating each token in sequence incrementally.",
|
613 |
"vocab": [
|
614 |
"a",
|
615 |
"b",
|
|
|
620 |
"files": [
|
621 |
{
|
622 |
"file_name": "edges.pkl",
|
623 |
+
"url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/26/edges.pkl"
|
624 |
},
|
625 |
{
|
626 |
+
"file_name": "ll_model.pth",
|
627 |
+
"url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/26/ll_model.pth"
|
628 |
},
|
629 |
{
|
630 |
+
"file_name": "ll_model_cfg.pkl",
|
631 |
+
"url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/26/ll_model_cfg.pkl"
|
632 |
},
|
633 |
{
|
634 |
+
"file_name": "meta.json",
|
635 |
+
"url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/26/meta.json"
|
636 |
}
|
637 |
],
|
638 |
"transformer_cfg": {
|
639 |
"n_layers": 2,
|
640 |
+
"d_model": 21,
|
641 |
"n_ctx": 10,
|
642 |
"d_head": 5,
|
643 |
"model_name": "custom",
|
644 |
"n_heads": 4,
|
645 |
+
"d_mlp": 84,
|
646 |
"act_fn": "gelu",
|
647 |
"d_vocab": 5,
|
648 |
"eps": 1e-05,
|
|
|
662 |
"attn_types": null,
|
663 |
"init_mode": "gpt2",
|
664 |
"normalization_type": null,
|
|
|
665 |
"n_devices": 1,
|
666 |
"attention_dir": "causal",
|
667 |
"attn_only": false,
|
668 |
"seed": 0,
|
669 |
+
"initializer_range": 0.12344267996967354,
|
670 |
"init_weights": true,
|
671 |
"scale_attn_by_inverse_layer_idx": false,
|
672 |
"positional_embedding_type": "standard",
|
673 |
"final_rms": false,
|
674 |
+
"d_vocab_out": 27,
|
675 |
"parallel_attn_mlp": false,
|
676 |
"rotary_dim": null,
|
677 |
+
"n_params": 10416,
|
678 |
"use_hook_tokens": false,
|
679 |
"gated_mlp": false,
|
680 |
"default_prepend_bos": true,
|
|
|
686 |
"trust_remote_code": false,
|
687 |
"rotary_adjacent_pairs": false
|
688 |
},
|
689 |
+
"transformer_cfg_file_url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/26/ll_model_cfg.pkl",
|
690 |
"training_args": {
|
691 |
"atol": 0.05,
|
692 |
"lr": 0.01,
|
|
|
694 |
"iit_weight": 1.0,
|
695 |
"behavior_weight": 1.0,
|
696 |
"strict_weight": 0.4,
|
697 |
+
"epochs": 2000,
|
698 |
"act_fn": "gelu",
|
699 |
+
"clip_grad_norm": 0.1,
|
700 |
+
"lr_scheduler": "",
|
701 |
+
"model_pair": "strict"
|
702 |
},
|
703 |
+
"training_args_file_url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/26/meta.json",
|
704 |
+
"weights_file_url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/26/ll_model.pth",
|
705 |
+
"circuit_file_url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/26/edges.pkl"
|
706 |
+
},
|
707 |
+
{
|
708 |
+
"case_id": "29",
|
709 |
+
"url": "https://huggingface.co/cybershiptrooper/InterpBench/tree/main/29",
|
710 |
+
"task_description": "Creates abbreviations for each token in the sequence.",
|
711 |
+
"vocab": [
|
712 |
+
"J",
|
713 |
+
"LB",
|
714 |
+
"TPSI",
|
715 |
+
"V",
|
716 |
+
"b",
|
717 |
+
"no",
|
718 |
+
"oCLrZaW",
|
719 |
+
"poiVg"
|
720 |
+
],
|
721 |
+
"max_seq_len": 10,
|
722 |
+
"min_seq_len": 4,
|
723 |
+
"files": [
|
724 |
+
{
|
725 |
+
"file_name": "edges.pkl",
|
726 |
+
"url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/29/edges.pkl"
|
727 |
+
},
|
728 |
+
{
|
729 |
+
"file_name": "ll_model.pth",
|
730 |
+
"url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/29/ll_model.pth"
|
731 |
+
},
|
732 |
+
{
|
733 |
+
"file_name": "ll_model_cfg.pkl",
|
734 |
+
"url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/29/ll_model_cfg.pkl"
|
735 |
+
},
|
736 |
+
{
|
737 |
+
"file_name": "meta.json",
|
738 |
+
"url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/29/meta.json"
|
739 |
+
}
|
740 |
+
],
|
741 |
+
"transformer_cfg": {
|
742 |
+
"n_layers": 2,
|
743 |
+
"d_model": 13,
|
744 |
+
"n_ctx": 10,
|
745 |
+
"d_head": 3,
|
746 |
+
"model_name": "custom",
|
747 |
+
"n_heads": 4,
|
748 |
+
"d_mlp": 52,
|
749 |
+
"act_fn": "gelu",
|
750 |
+
"d_vocab": 10,
|
751 |
+
"eps": 1e-05,
|
752 |
+
"use_attn_result": true,
|
753 |
+
"use_attn_scale": true,
|
754 |
+
"use_split_qkv_input": true,
|
755 |
+
"use_hook_mlp_in": true,
|
756 |
+
"use_attn_in": false,
|
757 |
+
"use_local_attn": false,
|
758 |
+
"original_architecture": null,
|
759 |
+
"from_checkpoint": false,
|
760 |
+
"checkpoint_index": null,
|
761 |
+
"checkpoint_label_type": null,
|
762 |
+
"checkpoint_value": null,
|
763 |
+
"tokenizer_name": null,
|
764 |
+
"window_size": null,
|
765 |
+
"attn_types": null,
|
766 |
+
"init_mode": "gpt2",
|
767 |
+
"normalization_type": null,
|
768 |
+
"n_devices": 1,
|
769 |
+
"attention_dir": "causal",
|
770 |
+
"attn_only": false,
|
771 |
+
"seed": 0,
|
772 |
+
"initializer_range": 0.1539600717839002,
|
773 |
+
"init_weights": true,
|
774 |
+
"scale_attn_by_inverse_layer_idx": false,
|
775 |
+
"positional_embedding_type": "standard",
|
776 |
+
"final_rms": false,
|
777 |
+
"d_vocab_out": 8,
|
778 |
+
"parallel_attn_mlp": false,
|
779 |
+
"rotary_dim": null,
|
780 |
+
"n_params": 3952,
|
781 |
+
"use_hook_tokens": false,
|
782 |
+
"gated_mlp": false,
|
783 |
+
"default_prepend_bos": true,
|
784 |
+
"dtype": "torch.float32",
|
785 |
+
"tokenizer_prepends_bos": null,
|
786 |
+
"n_key_value_heads": null,
|
787 |
+
"post_embedding_ln": false,
|
788 |
+
"rotary_base": 10000,
|
789 |
+
"trust_remote_code": false,
|
790 |
+
"rotary_adjacent_pairs": false
|
791 |
+
},
|
792 |
+
"transformer_cfg_file_url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/29/ll_model_cfg.pkl",
|
793 |
+
"training_args": {
|
794 |
+
"atol": 0.05,
|
795 |
+
"lr": 0.01,
|
796 |
+
"use_single_loss": false,
|
797 |
+
"iit_weight": 1.0,
|
798 |
+
"behavior_weight": 1.0,
|
799 |
+
"strict_weight": 0.4,
|
800 |
+
"epochs": 2000,
|
801 |
+
"act_fn": "gelu",
|
802 |
+
"clip_grad_norm": 0.1,
|
803 |
+
"lr_scheduler": "",
|
804 |
+
"model_pair": "strict"
|
805 |
+
},
|
806 |
+
"training_args_file_url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/29/meta.json",
|
807 |
+
"weights_file_url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/29/ll_model.pth",
|
808 |
+
"circuit_file_url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/29/edges.pkl"
|
809 |
},
|
810 |
{
|
811 |
"case_id": "3",
|
|
|
825 |
"url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/3/edges.pkl"
|
826 |
},
|
827 |
{
|
828 |
+
"file_name": "ll_model.pth",
|
829 |
+
"url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/3/ll_model.pth"
|
830 |
},
|
831 |
{
|
832 |
+
"file_name": "ll_model_cfg.pkl",
|
833 |
+
"url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/3/ll_model_cfg.pkl"
|
834 |
},
|
835 |
{
|
836 |
+
"file_name": "meta.json",
|
837 |
+
"url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/3/meta.json"
|
838 |
}
|
839 |
],
|
840 |
"transformer_cfg": {
|
|
|
864 |
"attn_types": null,
|
865 |
"init_mode": "gpt2",
|
866 |
"normalization_type": null,
|
|
|
867 |
"n_devices": 1,
|
868 |
"attention_dir": "causal",
|
869 |
"attn_only": false,
|
|
|
888 |
"trust_remote_code": false,
|
889 |
"rotary_adjacent_pairs": false
|
890 |
},
|
891 |
+
"transformer_cfg_file_url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/3/ll_model_cfg.pkl",
|
892 |
"training_args": {
|
893 |
"atol": 0.05,
|
894 |
"lr": 0.001,
|
|
|
901 |
"clip_grad_norm": 0.1,
|
902 |
"lr_scheduler": ""
|
903 |
},
|
904 |
+
"training_args_file_url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/3/meta.json",
|
905 |
+
"weights_file_url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/3/ll_model.pth",
|
906 |
"circuit_file_url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/3/edges.pkl"
|
907 |
},
|
908 |
{
|
|
|
927 |
"url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/33/edges.pkl"
|
928 |
},
|
929 |
{
|
930 |
+
"file_name": "ll_model.pth",
|
931 |
+
"url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/33/ll_model.pth"
|
932 |
},
|
933 |
{
|
934 |
+
"file_name": "ll_model_cfg.pkl",
|
935 |
+
"url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/33/ll_model_cfg.pkl"
|
936 |
},
|
937 |
{
|
938 |
+
"file_name": "meta.json",
|
939 |
+
"url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/33/meta.json"
|
940 |
}
|
941 |
],
|
942 |
"transformer_cfg": {
|
|
|
966 |
"attn_types": null,
|
967 |
"init_mode": "gpt2",
|
968 |
"normalization_type": null,
|
|
|
969 |
"n_devices": 1,
|
970 |
"attention_dir": "causal",
|
971 |
"attn_only": false,
|
|
|
990 |
"trust_remote_code": false,
|
991 |
"rotary_adjacent_pairs": false
|
992 |
},
|
993 |
+
"transformer_cfg_file_url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/33/ll_model_cfg.pkl",
|
994 |
"training_args": {
|
995 |
"atol": 0.05,
|
996 |
"lr": 0.001,
|
|
|
1003 |
"clip_grad_norm": 0.1,
|
1004 |
"lr_scheduler": ""
|
1005 |
},
|
1006 |
+
"training_args_file_url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/33/meta.json",
|
1007 |
+
"weights_file_url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/33/ll_model.pth",
|
1008 |
"circuit_file_url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/33/edges.pkl"
|
1009 |
},
|
1010 |
{
|
|
|
1029 |
"url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/34/edges.pkl"
|
1030 |
},
|
1031 |
{
|
1032 |
+
"file_name": "ll_model.pth",
|
1033 |
+
"url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/34/ll_model.pth"
|
1034 |
},
|
1035 |
{
|
1036 |
+
"file_name": "ll_model_cfg.pkl",
|
1037 |
+
"url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/34/ll_model_cfg.pkl"
|
1038 |
},
|
1039 |
{
|
1040 |
+
"file_name": "meta.json",
|
1041 |
+
"url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/34/meta.json"
|
1042 |
}
|
1043 |
],
|
1044 |
"transformer_cfg": {
|
1045 |
"n_layers": 2,
|
1046 |
+
"d_model": 16,
|
1047 |
"n_ctx": 10,
|
1048 |
+
"d_head": 4,
|
1049 |
"model_name": "custom",
|
1050 |
"n_heads": 4,
|
1051 |
+
"d_mlp": 64,
|
1052 |
"act_fn": "gelu",
|
1053 |
"d_vocab": 10,
|
1054 |
"eps": 1e-05,
|
|
|
1068 |
"attn_types": null,
|
1069 |
"init_mode": "gpt2",
|
1070 |
"normalization_type": null,
|
|
|
1071 |
"n_devices": 1,
|
1072 |
"attention_dir": "causal",
|
1073 |
"attn_only": false,
|
|
|
1080 |
"d_vocab_out": 5,
|
1081 |
"parallel_attn_mlp": false,
|
1082 |
"rotary_dim": null,
|
1083 |
+
"n_params": 6144,
|
1084 |
"use_hook_tokens": false,
|
1085 |
"gated_mlp": false,
|
1086 |
"default_prepend_bos": true,
|
|
|
1092 |
"trust_remote_code": false,
|
1093 |
"rotary_adjacent_pairs": false
|
1094 |
},
|
1095 |
+
"transformer_cfg_file_url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/34/ll_model_cfg.pkl",
|
1096 |
"training_args": {
|
1097 |
"atol": 0.05,
|
1098 |
+
"lr": 0.01,
|
1099 |
"use_single_loss": false,
|
1100 |
"iit_weight": 1.0,
|
1101 |
"behavior_weight": 1.0,
|
1102 |
+
"strict_weight": 1.0,
|
1103 |
"epochs": 2000,
|
1104 |
"act_fn": "gelu",
|
1105 |
"clip_grad_norm": 0.1,
|
1106 |
+
"lr_scheduler": "",
|
1107 |
+
"model_pair": "strict"
|
1108 |
},
|
1109 |
+
"training_args_file_url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/34/meta.json",
|
1110 |
+
"weights_file_url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/34/ll_model.pth",
|
1111 |
"circuit_file_url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/34/edges.pkl"
|
1112 |
},
|
1113 |
{
|
|
|
1132 |
"url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/35/edges.pkl"
|
1133 |
},
|
1134 |
{
|
1135 |
+
"file_name": "ll_model.pth",
|
1136 |
+
"url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/35/ll_model.pth"
|
1137 |
},
|
1138 |
{
|
1139 |
+
"file_name": "ll_model_cfg.pkl",
|
1140 |
+
"url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/35/ll_model_cfg.pkl"
|
1141 |
},
|
1142 |
{
|
1143 |
+
"file_name": "meta.json",
|
1144 |
+
"url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/35/meta.json"
|
1145 |
}
|
1146 |
],
|
1147 |
"transformer_cfg": {
|
1148 |
"n_layers": 2,
|
1149 |
+
"d_model": 9,
|
1150 |
"n_ctx": 10,
|
1151 |
+
"d_head": 2,
|
1152 |
"model_name": "custom",
|
1153 |
"n_heads": 4,
|
1154 |
+
"d_mlp": 36,
|
1155 |
"act_fn": "gelu",
|
1156 |
"d_vocab": 10,
|
1157 |
"eps": 1e-05,
|
|
|
1171 |
"attn_types": null,
|
1172 |
"init_mode": "gpt2",
|
1173 |
"normalization_type": null,
|
|
|
1174 |
"n_devices": 1,
|
1175 |
"attention_dir": "causal",
|
1176 |
"attn_only": false,
|
|
|
1183 |
"d_vocab_out": 8,
|
1184 |
"parallel_attn_mlp": false,
|
1185 |
"rotary_dim": null,
|
1186 |
+
"n_params": 1872,
|
1187 |
"use_hook_tokens": false,
|
1188 |
"gated_mlp": false,
|
1189 |
"default_prepend_bos": true,
|
|
|
1195 |
"trust_remote_code": false,
|
1196 |
"rotary_adjacent_pairs": false
|
1197 |
},
|
1198 |
+
"transformer_cfg_file_url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/35/ll_model_cfg.pkl",
|
1199 |
"training_args": {
|
1200 |
"atol": 0.05,
|
1201 |
+
"lr": 0.01,
|
1202 |
"use_single_loss": false,
|
1203 |
"iit_weight": 1.0,
|
1204 |
"behavior_weight": 1.0,
|
1205 |
+
"strict_weight": 1.0,
|
1206 |
"epochs": 2000,
|
1207 |
"act_fn": "gelu",
|
1208 |
"clip_grad_norm": 0.1,
|
1209 |
+
"lr_scheduler": "",
|
1210 |
+
"model_pair": "strict"
|
1211 |
},
|
1212 |
+
"training_args_file_url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/35/meta.json",
|
1213 |
+
"weights_file_url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/35/ll_model.pth",
|
1214 |
"circuit_file_url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/35/edges.pkl"
|
1215 |
},
|
1216 |
{
|
|
|
1230 |
"url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/36/edges.pkl"
|
1231 |
},
|
1232 |
{
|
1233 |
+
"file_name": "ll_model.pth",
|
1234 |
+
"url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/36/ll_model.pth"
|
1235 |
},
|
1236 |
{
|
1237 |
+
"file_name": "ll_model_cfg.pkl",
|
1238 |
+
"url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/36/ll_model_cfg.pkl"
|
1239 |
},
|
1240 |
{
|
1241 |
+
"file_name": "meta.json",
|
1242 |
+
"url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/36/meta.json"
|
1243 |
}
|
1244 |
],
|
1245 |
"transformer_cfg": {
|
1246 |
"n_layers": 2,
|
1247 |
+
"d_model": 6,
|
1248 |
"n_ctx": 10,
|
1249 |
"d_head": 1,
|
1250 |
"model_name": "custom",
|
1251 |
"n_heads": 4,
|
1252 |
+
"d_mlp": 24,
|
1253 |
"act_fn": "gelu",
|
1254 |
"d_vocab": 5,
|
1255 |
"eps": 1e-05,
|
|
|
1269 |
"attn_types": null,
|
1270 |
"init_mode": "gpt2",
|
1271 |
"normalization_type": null,
|
|
|
1272 |
"n_devices": 1,
|
1273 |
"attention_dir": "causal",
|
1274 |
"attn_only": false,
|
|
|
1281 |
"d_vocab_out": 3,
|
1282 |
"parallel_attn_mlp": false,
|
1283 |
"rotary_dim": null,
|
1284 |
+
"n_params": 768,
|
1285 |
"use_hook_tokens": false,
|
1286 |
"gated_mlp": false,
|
1287 |
"default_prepend_bos": true,
|
|
|
1293 |
"trust_remote_code": false,
|
1294 |
"rotary_adjacent_pairs": false
|
1295 |
},
|
1296 |
+
"transformer_cfg_file_url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/36/ll_model_cfg.pkl",
|
1297 |
"training_args": {
|
1298 |
"atol": 0.05,
|
1299 |
+
"lr": 0.01,
|
1300 |
"use_single_loss": false,
|
1301 |
"iit_weight": 1.0,
|
1302 |
"behavior_weight": 1.0,
|
1303 |
+
"strict_weight": 1.0,
|
1304 |
"epochs": 2000,
|
1305 |
"act_fn": "gelu",
|
1306 |
"clip_grad_norm": 0.1,
|
1307 |
+
"lr_scheduler": "",
|
1308 |
+
"model_pair": "strict"
|
1309 |
},
|
1310 |
+
"training_args_file_url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/36/meta.json",
|
1311 |
+
"weights_file_url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/36/ll_model.pth",
|
1312 |
"circuit_file_url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/36/edges.pkl"
|
1313 |
},
|
1314 |
{
|
|
|
1333 |
"url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/37/edges.pkl"
|
1334 |
},
|
1335 |
{
|
1336 |
+
"file_name": "ll_model.pth",
|
1337 |
+
"url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/37/ll_model.pth"
|
1338 |
},
|
1339 |
{
|
1340 |
+
"file_name": "ll_model_cfg.pkl",
|
1341 |
+
"url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/37/ll_model_cfg.pkl"
|
1342 |
},
|
1343 |
{
|
1344 |
+
"file_name": "meta.json",
|
1345 |
+
"url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/37/meta.json"
|
1346 |
}
|
1347 |
],
|
1348 |
"transformer_cfg": {
|
1349 |
"n_layers": 2,
|
1350 |
+
"d_model": 12,
|
1351 |
"n_ctx": 10,
|
1352 |
+
"d_head": 3,
|
1353 |
"model_name": "custom",
|
1354 |
"n_heads": 4,
|
1355 |
+
"d_mlp": 48,
|
1356 |
"act_fn": "gelu",
|
1357 |
"d_vocab": 10,
|
1358 |
"eps": 1e-05,
|
|
|
1372 |
"attn_types": null,
|
1373 |
"init_mode": "gpt2",
|
1374 |
"normalization_type": null,
|
|
|
1375 |
"n_devices": 1,
|
1376 |
"attention_dir": "causal",
|
1377 |
"attn_only": false,
|
|
|
1384 |
"d_vocab_out": 8,
|
1385 |
"parallel_attn_mlp": false,
|
1386 |
"rotary_dim": null,
|
1387 |
+
"n_params": 3456,
|
1388 |
"use_hook_tokens": false,
|
1389 |
"gated_mlp": false,
|
1390 |
"default_prepend_bos": true,
|
|
|
1396 |
"trust_remote_code": false,
|
1397 |
"rotary_adjacent_pairs": false
|
1398 |
},
|
1399 |
+
"transformer_cfg_file_url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/37/ll_model_cfg.pkl",
|
1400 |
"training_args": {
|
1401 |
"atol": 0.05,
|
1402 |
+
"lr": 0.01,
|
1403 |
"use_single_loss": false,
|
1404 |
"iit_weight": 1.0,
|
1405 |
"behavior_weight": 1.0,
|
1406 |
+
"strict_weight": 1.0,
|
1407 |
"epochs": 2000,
|
1408 |
"act_fn": "gelu",
|
1409 |
"clip_grad_norm": 0.1,
|
1410 |
+
"lr_scheduler": "",
|
1411 |
+
"model_pair": "strict"
|
1412 |
},
|
1413 |
+
"training_args_file_url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/37/meta.json",
|
1414 |
+
"weights_file_url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/37/ll_model.pth",
|
1415 |
"circuit_file_url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/37/edges.pkl"
|
1416 |
},
|
1417 |
{
|
1418 |
+
"case_id": "4",
|
1419 |
+
"url": "https://huggingface.co/cybershiptrooper/InterpBench/tree/main/4",
|
1420 |
+
"task_description": "Return fraction of previous open tokens minus the fraction of close tokens.",
|
1421 |
"vocab": [
|
1422 |
+
"(",
|
1423 |
+
")",
|
1424 |
"a",
|
1425 |
"b",
|
1426 |
"c"
|
|
|
1430 |
"files": [
|
1431 |
{
|
1432 |
"file_name": "edges.pkl",
|
1433 |
+
"url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/4/edges.pkl"
|
1434 |
},
|
1435 |
{
|
1436 |
+
"file_name": "ll_model.pth",
|
1437 |
+
"url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/4/ll_model.pth"
|
1438 |
},
|
1439 |
{
|
1440 |
+
"file_name": "ll_model_cfg.pkl",
|
1441 |
+
"url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/4/ll_model_cfg.pkl"
|
1442 |
},
|
1443 |
{
|
1444 |
+
"file_name": "meta.json",
|
1445 |
+
"url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/4/meta.json"
|
1446 |
}
|
1447 |
],
|
1448 |
"transformer_cfg": {
|
|
|
1454 |
"n_heads": 4,
|
1455 |
"d_mlp": 80,
|
1456 |
"act_fn": "gelu",
|
1457 |
+
"d_vocab": 7,
|
1458 |
"eps": 1e-05,
|
1459 |
"use_attn_result": true,
|
1460 |
"use_attn_scale": true,
|
|
|
1472 |
"attn_types": null,
|
1473 |
"init_mode": "gpt2",
|
1474 |
"normalization_type": null,
|
|
|
1475 |
"n_devices": 1,
|
1476 |
"attention_dir": "causal",
|
1477 |
"attn_only": false,
|
1478 |
"seed": 0,
|
1479 |
+
"initializer_range": 0.17056057308448835,
|
1480 |
"init_weights": true,
|
1481 |
"scale_attn_by_inverse_layer_idx": false,
|
1482 |
"positional_embedding_type": "standard",
|
1483 |
"final_rms": false,
|
1484 |
+
"d_vocab_out": 1,
|
1485 |
"parallel_attn_mlp": false,
|
1486 |
"rotary_dim": null,
|
1487 |
"n_params": 9600,
|
|
|
1496 |
"trust_remote_code": false,
|
1497 |
"rotary_adjacent_pairs": false
|
1498 |
},
|
1499 |
+
"transformer_cfg_file_url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/4/ll_model_cfg.pkl",
|
1500 |
"training_args": {
|
1501 |
"atol": 0.05,
|
1502 |
"lr": 0.001,
|
|
|
1509 |
"clip_grad_norm": 0.1,
|
1510 |
"lr_scheduler": ""
|
1511 |
},
|
1512 |
+
"training_args_file_url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/4/meta.json",
|
1513 |
+
"weights_file_url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/4/ll_model.pth",
|
1514 |
+
"circuit_file_url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/4/edges.pkl"
|
1515 |
},
|
1516 |
{
|
1517 |
+
"case_id": "8",
|
1518 |
+
"url": "https://huggingface.co/cybershiptrooper/InterpBench/tree/main/8",
|
1519 |
+
"task_description": "Fills gaps between tokens with a specified filler.",
|
1520 |
"vocab": [
|
1521 |
+
"J",
|
1522 |
+
"LB",
|
1523 |
+
"TPSI",
|
1524 |
+
"V",
|
1525 |
"b",
|
1526 |
+
"no",
|
1527 |
+
"oCLrZaW",
|
1528 |
+
"poiVg"
|
1529 |
],
|
1530 |
"max_seq_len": 10,
|
1531 |
"min_seq_len": 4,
|
1532 |
"files": [
|
1533 |
{
|
1534 |
"file_name": "edges.pkl",
|
1535 |
+
"url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/8/edges.pkl"
|
1536 |
},
|
1537 |
{
|
1538 |
+
"file_name": "ll_model.pth",
|
1539 |
+
"url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/8/ll_model.pth"
|
1540 |
},
|
1541 |
{
|
1542 |
+
"file_name": "ll_model_cfg.pkl",
|
1543 |
+
"url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/8/ll_model_cfg.pkl"
|
1544 |
},
|
1545 |
{
|
1546 |
+
"file_name": "meta.json",
|
1547 |
+
"url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/8/meta.json"
|
1548 |
}
|
1549 |
],
|
1550 |
"transformer_cfg": {
|
|
|
1556 |
"n_heads": 4,
|
1557 |
"d_mlp": 80,
|
1558 |
"act_fn": "gelu",
|
1559 |
+
"d_vocab": 10,
|
1560 |
"eps": 1e-05,
|
1561 |
"use_attn_result": true,
|
1562 |
"use_attn_scale": true,
|
|
|
1574 |
"attn_types": null,
|
1575 |
"init_mode": "gpt2",
|
1576 |
"normalization_type": null,
|
|
|
1577 |
"n_devices": 1,
|
1578 |
"attention_dir": "causal",
|
1579 |
"attn_only": false,
|
1580 |
"seed": 0,
|
1581 |
+
"initializer_range": 0.13333333333333333,
|
1582 |
"init_weights": true,
|
1583 |
"scale_attn_by_inverse_layer_idx": false,
|
1584 |
"positional_embedding_type": "standard",
|
1585 |
"final_rms": false,
|
1586 |
+
"d_vocab_out": 8,
|
1587 |
"parallel_attn_mlp": false,
|
1588 |
"rotary_dim": null,
|
1589 |
"n_params": 9600,
|
|
|
1598 |
"trust_remote_code": false,
|
1599 |
"rotary_adjacent_pairs": false
|
1600 |
},
|
1601 |
+
"transformer_cfg_file_url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/8/ll_model_cfg.pkl",
|
1602 |
"training_args": {
|
1603 |
"atol": 0.05,
|
1604 |
+
"lr": 0.01,
|
1605 |
"use_single_loss": false,
|
1606 |
"iit_weight": 1.0,
|
1607 |
"behavior_weight": 1.0,
|
1608 |
"strict_weight": 0.4,
|
1609 |
+
"epochs": 500,
|
1610 |
"act_fn": "gelu",
|
1611 |
+
"clip_grad_norm": 1.0,
|
1612 |
"lr_scheduler": ""
|
1613 |
},
|
1614 |
+
"training_args_file_url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/8/meta.json",
|
1615 |
+
"weights_file_url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/8/ll_model.pth",
|
1616 |
+
"circuit_file_url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/8/edges.pkl"
|
1617 |
},
|
1618 |
{
|
1619 |
+
"case_id": "ioi",
|
1620 |
+
"url": "https://huggingface.co/cybershiptrooper/InterpBench/tree/main/ioi",
|
1621 |
+
"task_description": "Indirect object identification",
|
1622 |
+
"max_seq_len": 16,
|
1623 |
+
"min_seq_len": 16,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1624 |
"files": [
|
1625 |
{
|
1626 |
"file_name": "edges.pkl",
|
1627 |
+
"url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/ioi/edges.pkl"
|
1628 |
},
|
1629 |
{
|
1630 |
+
"file_name": "ll_model.pth",
|
1631 |
+
"url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/ioi/ll_model.pth"
|
1632 |
},
|
1633 |
{
|
1634 |
+
"file_name": "ll_model_cfg.pkl",
|
1635 |
+
"url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/ioi/ll_model_cfg.pkl"
|
1636 |
},
|
1637 |
{
|
1638 |
+
"file_name": "meta.json",
|
1639 |
+
"url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/ioi/meta.json"
|
1640 |
}
|
1641 |
],
|
1642 |
"transformer_cfg": {
|
1643 |
+
"n_layers": 6,
|
1644 |
+
"d_model": 64,
|
1645 |
+
"n_ctx": 1024,
|
1646 |
+
"d_head": 16,
|
1647 |
+
"model_name": "gpt2",
|
1648 |
"n_heads": 4,
|
1649 |
+
"d_mlp": 3072,
|
1650 |
+
"act_fn": "gelu_new",
|
1651 |
+
"d_vocab": 50257,
|
1652 |
"eps": 1e-05,
|
1653 |
+
"use_attn_result": false,
|
1654 |
"use_attn_scale": true,
|
1655 |
+
"use_split_qkv_input": false,
|
1656 |
+
"use_hook_mlp_in": false,
|
1657 |
"use_attn_in": false,
|
1658 |
"use_local_attn": false,
|
1659 |
+
"original_architecture": "GPT2LMHeadModel",
|
1660 |
"from_checkpoint": false,
|
1661 |
"checkpoint_index": null,
|
1662 |
"checkpoint_label_type": null,
|
1663 |
"checkpoint_value": null,
|
1664 |
+
"tokenizer_name": "gpt2",
|
1665 |
"window_size": null,
|
1666 |
"attn_types": null,
|
1667 |
"init_mode": "gpt2",
|
1668 |
+
"normalization_type": "LNPre",
|
|
|
1669 |
"n_devices": 1,
|
1670 |
"attention_dir": "causal",
|
1671 |
"attn_only": false,
|
1672 |
+
"seed": null,
|
1673 |
+
"initializer_range": 0.02886751345948129,
|
1674 |
"init_weights": true,
|
1675 |
"scale_attn_by_inverse_layer_idx": false,
|
1676 |
"positional_embedding_type": "standard",
|
1677 |
"final_rms": false,
|
1678 |
+
"d_vocab_out": 50257,
|
1679 |
"parallel_attn_mlp": false,
|
1680 |
"rotary_dim": null,
|
1681 |
+
"n_params": 2457600,
|
1682 |
"use_hook_tokens": false,
|
1683 |
"gated_mlp": false,
|
1684 |
"default_prepend_bos": true,
|
1685 |
"dtype": "torch.float32",
|
1686 |
+
"tokenizer_prepends_bos": false,
|
1687 |
"n_key_value_heads": null,
|
1688 |
"post_embedding_ln": false,
|
1689 |
"rotary_base": 10000,
|
1690 |
"trust_remote_code": false,
|
1691 |
"rotary_adjacent_pairs": false
|
1692 |
},
|
1693 |
+
"transformer_cfg_file_url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/ioi/ll_model_cfg.pkl",
|
1694 |
"training_args": {
|
1695 |
+
"next_token": true,
|
1696 |
+
"non_ioi_thresh": 0.65,
|
1697 |
+
"use_per_token_check": false,
|
1698 |
+
"batch_size": 256,
|
1699 |
+
"lr": 0.001,
|
1700 |
+
"num_workers": 0,
|
1701 |
+
"early_stop": true,
|
1702 |
+
"lr_scheduler": null,
|
1703 |
+
"scheduler_val_metric": [
|
1704 |
+
"val/accuracy",
|
1705 |
+
"val/IIA"
|
1706 |
+
],
|
1707 |
+
"scheduler_mode": "max",
|
1708 |
+
"clip_grad_norm": 1.0,
|
1709 |
"atol": 0.05,
|
|
|
1710 |
"use_single_loss": false,
|
1711 |
"iit_weight": 1.0,
|
1712 |
"behavior_weight": 1.0,
|
1713 |
+
"strict_weight": 0.4
|
|
|
|
|
|
|
|
|
1714 |
},
|
1715 |
+
"training_args_file_url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/ioi/meta.json",
|
1716 |
+
"weights_file_url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/ioi/ll_model.pth",
|
1717 |
+
"circuit_file_url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/ioi/edges.pkl"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1718 |
},
|
1719 |
{
|
1720 |
"case_id": "ioi_next_token",
|
|
|
1724 |
"min_seq_len": 16,
|
1725 |
"files": [
|
1726 |
{
|
1727 |
+
"file_name": "edges.pkl",
|
1728 |
+
"url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/ioi_next_token/edges.pkl"
|
1729 |
+
},
|
1730 |
+
{
|
1731 |
+
"file_name": "ll_model.pth",
|
1732 |
+
"url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/ioi_next_token/ll_model.pth"
|
1733 |
},
|
1734 |
{
|
1735 |
+
"file_name": "ll_model_cfg.pkl",
|
1736 |
+
"url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/ioi_next_token/ll_model_cfg.pkl"
|
1737 |
},
|
1738 |
{
|
1739 |
+
"file_name": "meta.json",
|
1740 |
+
"url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/ioi_next_token/meta.json"
|
1741 |
}
|
1742 |
],
|
1743 |
+
"transformer_cfg": {
|
1744 |
+
"n_layers": 6,
|
1745 |
+
"d_model": 64,
|
1746 |
+
"n_ctx": 1024,
|
1747 |
+
"d_head": 16,
|
1748 |
+
"model_name": "gpt2",
|
1749 |
+
"n_heads": 4,
|
1750 |
+
"d_mlp": 3072,
|
1751 |
+
"act_fn": "gelu_new",
|
1752 |
+
"d_vocab": 50257,
|
1753 |
+
"eps": 1e-05,
|
1754 |
+
"use_attn_result": false,
|
1755 |
+
"use_attn_scale": true,
|
1756 |
+
"use_split_qkv_input": false,
|
1757 |
+
"use_hook_mlp_in": false,
|
1758 |
+
"use_attn_in": false,
|
1759 |
+
"use_local_attn": false,
|
1760 |
+
"original_architecture": "GPT2LMHeadModel",
|
1761 |
+
"from_checkpoint": false,
|
1762 |
+
"checkpoint_index": null,
|
1763 |
+
"checkpoint_label_type": null,
|
1764 |
+
"checkpoint_value": null,
|
1765 |
+
"tokenizer_name": "gpt2",
|
1766 |
+
"window_size": null,
|
1767 |
+
"attn_types": null,
|
1768 |
+
"init_mode": "gpt2",
|
1769 |
+
"normalization_type": "LNPre",
|
1770 |
+
"n_devices": 1,
|
1771 |
+
"attention_dir": "causal",
|
1772 |
+
"attn_only": false,
|
1773 |
+
"seed": null,
|
1774 |
+
"initializer_range": 0.02886751345948129,
|
1775 |
+
"init_weights": true,
|
1776 |
+
"scale_attn_by_inverse_layer_idx": false,
|
1777 |
+
"positional_embedding_type": "standard",
|
1778 |
+
"final_rms": false,
|
1779 |
+
"d_vocab_out": 50257,
|
1780 |
+
"parallel_attn_mlp": false,
|
1781 |
+
"rotary_dim": null,
|
1782 |
+
"n_params": 2457600,
|
1783 |
+
"use_hook_tokens": false,
|
1784 |
+
"gated_mlp": false,
|
1785 |
+
"default_prepend_bos": true,
|
1786 |
+
"dtype": "torch.float32",
|
1787 |
+
"tokenizer_prepends_bos": false,
|
1788 |
+
"n_key_value_heads": null,
|
1789 |
+
"post_embedding_ln": false,
|
1790 |
+
"rotary_base": 10000,
|
1791 |
+
"trust_remote_code": false,
|
1792 |
+
"rotary_adjacent_pairs": false
|
1793 |
+
},
|
1794 |
+
"transformer_cfg_file_url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/ioi_next_token/ll_model_cfg.pkl",
|
1795 |
+
"training_args": {
|
1796 |
+
"next_token": true,
|
1797 |
+
"non_ioi_thresh": 0.65,
|
1798 |
+
"use_per_token_check": false,
|
1799 |
+
"batch_size": 256,
|
1800 |
+
"lr": 0.001,
|
1801 |
+
"num_workers": 0,
|
1802 |
+
"early_stop": true,
|
1803 |
+
"lr_scheduler": null,
|
1804 |
+
"scheduler_val_metric": [
|
1805 |
+
"val/accuracy",
|
1806 |
+
"val/IIA"
|
1807 |
+
],
|
1808 |
+
"scheduler_mode": "max",
|
1809 |
+
"clip_grad_norm": 1.0,
|
1810 |
+
"atol": 0.05,
|
1811 |
+
"use_single_loss": false,
|
1812 |
+
"iit_weight": 1.0,
|
1813 |
+
"behavior_weight": 1.0,
|
1814 |
+
"strict_weight": 0.4
|
1815 |
+
},
|
1816 |
+
"training_args_file_url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/ioi_next_token/meta.json",
|
1817 |
+
"weights_file_url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/ioi_next_token/ll_model.pth",
|
1818 |
+
"circuit_file_url": "https://huggingface.co/cybershiptrooper/InterpBench/blob/main/ioi_next_token/edges.pkl"
|
1819 |
}
|
1820 |
]
|
1821 |
}
|
benchmark_metadata_croissant.json
CHANGED
@@ -412,7 +412,7 @@
|
|
412 |
"@id": "transformer_cfg.n_layers",
|
413 |
"name": "transformer_cfg.n_layers",
|
414 |
"description": "Column 'transformer_cfg.n_layers' from the parquet file describing all the cases in the benchmark.",
|
415 |
-
"dataType": "sc:
|
416 |
"source": {
|
417 |
"fileSet": {
|
418 |
"@id": "benchmark-cases-parquet"
|
@@ -427,7 +427,7 @@
|
|
427 |
"@id": "transformer_cfg.d_model",
|
428 |
"name": "transformer_cfg.d_model",
|
429 |
"description": "Column 'transformer_cfg.d_model' from the parquet file describing all the cases in the benchmark.",
|
430 |
-
"dataType": "sc:
|
431 |
"source": {
|
432 |
"fileSet": {
|
433 |
"@id": "benchmark-cases-parquet"
|
@@ -442,7 +442,7 @@
|
|
442 |
"@id": "transformer_cfg.n_ctx",
|
443 |
"name": "transformer_cfg.n_ctx",
|
444 |
"description": "Column 'transformer_cfg.n_ctx' from the parquet file describing all the cases in the benchmark.",
|
445 |
-
"dataType": "sc:
|
446 |
"source": {
|
447 |
"fileSet": {
|
448 |
"@id": "benchmark-cases-parquet"
|
@@ -457,7 +457,7 @@
|
|
457 |
"@id": "transformer_cfg.d_head",
|
458 |
"name": "transformer_cfg.d_head",
|
459 |
"description": "Column 'transformer_cfg.d_head' from the parquet file describing all the cases in the benchmark.",
|
460 |
-
"dataType": "sc:
|
461 |
"source": {
|
462 |
"fileSet": {
|
463 |
"@id": "benchmark-cases-parquet"
|
@@ -487,7 +487,7 @@
|
|
487 |
"@id": "transformer_cfg.n_heads",
|
488 |
"name": "transformer_cfg.n_heads",
|
489 |
"description": "Column 'transformer_cfg.n_heads' from the parquet file describing all the cases in the benchmark.",
|
490 |
-
"dataType": "sc:
|
491 |
"source": {
|
492 |
"fileSet": {
|
493 |
"@id": "benchmark-cases-parquet"
|
@@ -502,7 +502,7 @@
|
|
502 |
"@id": "transformer_cfg.d_mlp",
|
503 |
"name": "transformer_cfg.d_mlp",
|
504 |
"description": "Column 'transformer_cfg.d_mlp' from the parquet file describing all the cases in the benchmark.",
|
505 |
-
"dataType": "sc:
|
506 |
"source": {
|
507 |
"fileSet": {
|
508 |
"@id": "benchmark-cases-parquet"
|
@@ -532,7 +532,7 @@
|
|
532 |
"@id": "transformer_cfg.d_vocab",
|
533 |
"name": "transformer_cfg.d_vocab",
|
534 |
"description": "Column 'transformer_cfg.d_vocab' from the parquet file describing all the cases in the benchmark.",
|
535 |
-
"dataType": "sc:
|
536 |
"source": {
|
537 |
"fileSet": {
|
538 |
"@id": "benchmark-cases-parquet"
|
@@ -652,7 +652,7 @@
|
|
652 |
"@id": "transformer_cfg.original_architecture",
|
653 |
"name": "transformer_cfg.original_architecture",
|
654 |
"description": "Column 'transformer_cfg.original_architecture' from the parquet file describing all the cases in the benchmark.",
|
655 |
-
"dataType": "sc:
|
656 |
"source": {
|
657 |
"fileSet": {
|
658 |
"@id": "benchmark-cases-parquet"
|
@@ -677,57 +677,12 @@
|
|
677 |
}
|
678 |
}
|
679 |
},
|
680 |
-
{
|
681 |
-
"@type": "cr:Field",
|
682 |
-
"@id": "transformer_cfg.checkpoint_index",
|
683 |
-
"name": "transformer_cfg.checkpoint_index",
|
684 |
-
"description": "Column 'transformer_cfg.checkpoint_index' from the parquet file describing all the cases in the benchmark.",
|
685 |
-
"dataType": "sc:Float",
|
686 |
-
"source": {
|
687 |
-
"fileSet": {
|
688 |
-
"@id": "benchmark-cases-parquet"
|
689 |
-
},
|
690 |
-
"extract": {
|
691 |
-
"column": "transformer_cfg.checkpoint_index"
|
692 |
-
}
|
693 |
-
}
|
694 |
-
},
|
695 |
-
{
|
696 |
-
"@type": "cr:Field",
|
697 |
-
"@id": "transformer_cfg.checkpoint_label_type",
|
698 |
-
"name": "transformer_cfg.checkpoint_label_type",
|
699 |
-
"description": "Column 'transformer_cfg.checkpoint_label_type' from the parquet file describing all the cases in the benchmark.",
|
700 |
-
"dataType": "sc:Float",
|
701 |
-
"source": {
|
702 |
-
"fileSet": {
|
703 |
-
"@id": "benchmark-cases-parquet"
|
704 |
-
},
|
705 |
-
"extract": {
|
706 |
-
"column": "transformer_cfg.checkpoint_label_type"
|
707 |
-
}
|
708 |
-
}
|
709 |
-
},
|
710 |
-
{
|
711 |
-
"@type": "cr:Field",
|
712 |
-
"@id": "transformer_cfg.checkpoint_value",
|
713 |
-
"name": "transformer_cfg.checkpoint_value",
|
714 |
-
"description": "Column 'transformer_cfg.checkpoint_value' from the parquet file describing all the cases in the benchmark.",
|
715 |
-
"dataType": "sc:Float",
|
716 |
-
"source": {
|
717 |
-
"fileSet": {
|
718 |
-
"@id": "benchmark-cases-parquet"
|
719 |
-
},
|
720 |
-
"extract": {
|
721 |
-
"column": "transformer_cfg.checkpoint_value"
|
722 |
-
}
|
723 |
-
}
|
724 |
-
},
|
725 |
{
|
726 |
"@type": "cr:Field",
|
727 |
"@id": "transformer_cfg.tokenizer_name",
|
728 |
"name": "transformer_cfg.tokenizer_name",
|
729 |
"description": "Column 'transformer_cfg.tokenizer_name' from the parquet file describing all the cases in the benchmark.",
|
730 |
-
"dataType": "sc:
|
731 |
"source": {
|
732 |
"fileSet": {
|
733 |
"@id": "benchmark-cases-parquet"
|
@@ -737,36 +692,6 @@
|
|
737 |
}
|
738 |
}
|
739 |
},
|
740 |
-
{
|
741 |
-
"@type": "cr:Field",
|
742 |
-
"@id": "transformer_cfg.window_size",
|
743 |
-
"name": "transformer_cfg.window_size",
|
744 |
-
"description": "Column 'transformer_cfg.window_size' from the parquet file describing all the cases in the benchmark.",
|
745 |
-
"dataType": "sc:Float",
|
746 |
-
"source": {
|
747 |
-
"fileSet": {
|
748 |
-
"@id": "benchmark-cases-parquet"
|
749 |
-
},
|
750 |
-
"extract": {
|
751 |
-
"column": "transformer_cfg.window_size"
|
752 |
-
}
|
753 |
-
}
|
754 |
-
},
|
755 |
-
{
|
756 |
-
"@type": "cr:Field",
|
757 |
-
"@id": "transformer_cfg.attn_types",
|
758 |
-
"name": "transformer_cfg.attn_types",
|
759 |
-
"description": "Column 'transformer_cfg.attn_types' from the parquet file describing all the cases in the benchmark.",
|
760 |
-
"dataType": "sc:Float",
|
761 |
-
"source": {
|
762 |
-
"fileSet": {
|
763 |
-
"@id": "benchmark-cases-parquet"
|
764 |
-
},
|
765 |
-
"extract": {
|
766 |
-
"column": "transformer_cfg.attn_types"
|
767 |
-
}
|
768 |
-
}
|
769 |
-
},
|
770 |
{
|
771 |
"@type": "cr:Field",
|
772 |
"@id": "transformer_cfg.init_mode",
|
@@ -787,28 +712,13 @@
|
|
787 |
"@id": "transformer_cfg.normalization_type",
|
788 |
"name": "transformer_cfg.normalization_type",
|
789 |
"description": "Column 'transformer_cfg.normalization_type' from the parquet file describing all the cases in the benchmark.",
|
790 |
-
"dataType": "sc:Float",
|
791 |
-
"source": {
|
792 |
-
"fileSet": {
|
793 |
-
"@id": "benchmark-cases-parquet"
|
794 |
-
},
|
795 |
-
"extract": {
|
796 |
-
"column": "transformer_cfg.normalization_type"
|
797 |
-
}
|
798 |
-
}
|
799 |
-
},
|
800 |
-
{
|
801 |
-
"@type": "cr:Field",
|
802 |
-
"@id": "transformer_cfg.device",
|
803 |
-
"name": "transformer_cfg.device",
|
804 |
-
"description": "Column 'transformer_cfg.device' from the parquet file describing all the cases in the benchmark.",
|
805 |
"dataType": "sc:Text",
|
806 |
"source": {
|
807 |
"fileSet": {
|
808 |
"@id": "benchmark-cases-parquet"
|
809 |
},
|
810 |
"extract": {
|
811 |
-
"column": "transformer_cfg.
|
812 |
}
|
813 |
}
|
814 |
},
|
@@ -817,7 +727,7 @@
|
|
817 |
"@id": "transformer_cfg.n_devices",
|
818 |
"name": "transformer_cfg.n_devices",
|
819 |
"description": "Column 'transformer_cfg.n_devices' from the parquet file describing all the cases in the benchmark.",
|
820 |
-
"dataType": "sc:
|
821 |
"source": {
|
822 |
"fileSet": {
|
823 |
"@id": "benchmark-cases-parquet"
|
@@ -952,7 +862,7 @@
|
|
952 |
"@id": "transformer_cfg.d_vocab_out",
|
953 |
"name": "transformer_cfg.d_vocab_out",
|
954 |
"description": "Column 'transformer_cfg.d_vocab_out' from the parquet file describing all the cases in the benchmark.",
|
955 |
-
"dataType": "sc:
|
956 |
"source": {
|
957 |
"fileSet": {
|
958 |
"@id": "benchmark-cases-parquet"
|
@@ -977,27 +887,12 @@
|
|
977 |
}
|
978 |
}
|
979 |
},
|
980 |
-
{
|
981 |
-
"@type": "cr:Field",
|
982 |
-
"@id": "transformer_cfg.rotary_dim",
|
983 |
-
"name": "transformer_cfg.rotary_dim",
|
984 |
-
"description": "Column 'transformer_cfg.rotary_dim' from the parquet file describing all the cases in the benchmark.",
|
985 |
-
"dataType": "sc:Float",
|
986 |
-
"source": {
|
987 |
-
"fileSet": {
|
988 |
-
"@id": "benchmark-cases-parquet"
|
989 |
-
},
|
990 |
-
"extract": {
|
991 |
-
"column": "transformer_cfg.rotary_dim"
|
992 |
-
}
|
993 |
-
}
|
994 |
-
},
|
995 |
{
|
996 |
"@type": "cr:Field",
|
997 |
"@id": "transformer_cfg.n_params",
|
998 |
"name": "transformer_cfg.n_params",
|
999 |
"description": "Column 'transformer_cfg.n_params' from the parquet file describing all the cases in the benchmark.",
|
1000 |
-
"dataType": "sc:
|
1001 |
"source": {
|
1002 |
"fileSet": {
|
1003 |
"@id": "benchmark-cases-parquet"
|
@@ -1072,7 +967,7 @@
|
|
1072 |
"@id": "transformer_cfg.tokenizer_prepends_bos",
|
1073 |
"name": "transformer_cfg.tokenizer_prepends_bos",
|
1074 |
"description": "Column 'transformer_cfg.tokenizer_prepends_bos' from the parquet file describing all the cases in the benchmark.",
|
1075 |
-
"dataType": "sc:
|
1076 |
"source": {
|
1077 |
"fileSet": {
|
1078 |
"@id": "benchmark-cases-parquet"
|
@@ -1082,21 +977,6 @@
|
|
1082 |
}
|
1083 |
}
|
1084 |
},
|
1085 |
-
{
|
1086 |
-
"@type": "cr:Field",
|
1087 |
-
"@id": "transformer_cfg.n_key_value_heads",
|
1088 |
-
"name": "transformer_cfg.n_key_value_heads",
|
1089 |
-
"description": "Column 'transformer_cfg.n_key_value_heads' from the parquet file describing all the cases in the benchmark.",
|
1090 |
-
"dataType": "sc:Float",
|
1091 |
-
"source": {
|
1092 |
-
"fileSet": {
|
1093 |
-
"@id": "benchmark-cases-parquet"
|
1094 |
-
},
|
1095 |
-
"extract": {
|
1096 |
-
"column": "transformer_cfg.n_key_value_heads"
|
1097 |
-
}
|
1098 |
-
}
|
1099 |
-
},
|
1100 |
{
|
1101 |
"@type": "cr:Field",
|
1102 |
"@id": "transformer_cfg.post_embedding_ln",
|
@@ -1117,7 +997,7 @@
|
|
1117 |
"@id": "transformer_cfg.rotary_base",
|
1118 |
"name": "transformer_cfg.rotary_base",
|
1119 |
"description": "Column 'transformer_cfg.rotary_base' from the parquet file describing all the cases in the benchmark.",
|
1120 |
-
"dataType": "sc:
|
1121 |
"source": {
|
1122 |
"fileSet": {
|
1123 |
"@id": "benchmark-cases-parquet"
|
@@ -1156,6 +1036,141 @@
|
|
1156 |
"column": "transformer_cfg.rotary_adjacent_pairs"
|
1157 |
}
|
1158 |
}
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1159 |
}
|
1160 |
]
|
1161 |
}
|
|
|
412 |
"@id": "transformer_cfg.n_layers",
|
413 |
"name": "transformer_cfg.n_layers",
|
414 |
"description": "Column 'transformer_cfg.n_layers' from the parquet file describing all the cases in the benchmark.",
|
415 |
+
"dataType": "sc:Integer",
|
416 |
"source": {
|
417 |
"fileSet": {
|
418 |
"@id": "benchmark-cases-parquet"
|
|
|
427 |
"@id": "transformer_cfg.d_model",
|
428 |
"name": "transformer_cfg.d_model",
|
429 |
"description": "Column 'transformer_cfg.d_model' from the parquet file describing all the cases in the benchmark.",
|
430 |
+
"dataType": "sc:Integer",
|
431 |
"source": {
|
432 |
"fileSet": {
|
433 |
"@id": "benchmark-cases-parquet"
|
|
|
442 |
"@id": "transformer_cfg.n_ctx",
|
443 |
"name": "transformer_cfg.n_ctx",
|
444 |
"description": "Column 'transformer_cfg.n_ctx' from the parquet file describing all the cases in the benchmark.",
|
445 |
+
"dataType": "sc:Integer",
|
446 |
"source": {
|
447 |
"fileSet": {
|
448 |
"@id": "benchmark-cases-parquet"
|
|
|
457 |
"@id": "transformer_cfg.d_head",
|
458 |
"name": "transformer_cfg.d_head",
|
459 |
"description": "Column 'transformer_cfg.d_head' from the parquet file describing all the cases in the benchmark.",
|
460 |
+
"dataType": "sc:Integer",
|
461 |
"source": {
|
462 |
"fileSet": {
|
463 |
"@id": "benchmark-cases-parquet"
|
|
|
487 |
"@id": "transformer_cfg.n_heads",
|
488 |
"name": "transformer_cfg.n_heads",
|
489 |
"description": "Column 'transformer_cfg.n_heads' from the parquet file describing all the cases in the benchmark.",
|
490 |
+
"dataType": "sc:Integer",
|
491 |
"source": {
|
492 |
"fileSet": {
|
493 |
"@id": "benchmark-cases-parquet"
|
|
|
502 |
"@id": "transformer_cfg.d_mlp",
|
503 |
"name": "transformer_cfg.d_mlp",
|
504 |
"description": "Column 'transformer_cfg.d_mlp' from the parquet file describing all the cases in the benchmark.",
|
505 |
+
"dataType": "sc:Integer",
|
506 |
"source": {
|
507 |
"fileSet": {
|
508 |
"@id": "benchmark-cases-parquet"
|
|
|
532 |
"@id": "transformer_cfg.d_vocab",
|
533 |
"name": "transformer_cfg.d_vocab",
|
534 |
"description": "Column 'transformer_cfg.d_vocab' from the parquet file describing all the cases in the benchmark.",
|
535 |
+
"dataType": "sc:Integer",
|
536 |
"source": {
|
537 |
"fileSet": {
|
538 |
"@id": "benchmark-cases-parquet"
|
|
|
652 |
"@id": "transformer_cfg.original_architecture",
|
653 |
"name": "transformer_cfg.original_architecture",
|
654 |
"description": "Column 'transformer_cfg.original_architecture' from the parquet file describing all the cases in the benchmark.",
|
655 |
+
"dataType": "sc:Text",
|
656 |
"source": {
|
657 |
"fileSet": {
|
658 |
"@id": "benchmark-cases-parquet"
|
|
|
677 |
}
|
678 |
}
|
679 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
680 |
{
|
681 |
"@type": "cr:Field",
|
682 |
"@id": "transformer_cfg.tokenizer_name",
|
683 |
"name": "transformer_cfg.tokenizer_name",
|
684 |
"description": "Column 'transformer_cfg.tokenizer_name' from the parquet file describing all the cases in the benchmark.",
|
685 |
+
"dataType": "sc:Text",
|
686 |
"source": {
|
687 |
"fileSet": {
|
688 |
"@id": "benchmark-cases-parquet"
|
|
|
692 |
}
|
693 |
}
|
694 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
695 |
{
|
696 |
"@type": "cr:Field",
|
697 |
"@id": "transformer_cfg.init_mode",
|
|
|
712 |
"@id": "transformer_cfg.normalization_type",
|
713 |
"name": "transformer_cfg.normalization_type",
|
714 |
"description": "Column 'transformer_cfg.normalization_type' from the parquet file describing all the cases in the benchmark.",
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
715 |
"dataType": "sc:Text",
|
716 |
"source": {
|
717 |
"fileSet": {
|
718 |
"@id": "benchmark-cases-parquet"
|
719 |
},
|
720 |
"extract": {
|
721 |
+
"column": "transformer_cfg.normalization_type"
|
722 |
}
|
723 |
}
|
724 |
},
|
|
|
727 |
"@id": "transformer_cfg.n_devices",
|
728 |
"name": "transformer_cfg.n_devices",
|
729 |
"description": "Column 'transformer_cfg.n_devices' from the parquet file describing all the cases in the benchmark.",
|
730 |
+
"dataType": "sc:Integer",
|
731 |
"source": {
|
732 |
"fileSet": {
|
733 |
"@id": "benchmark-cases-parquet"
|
|
|
862 |
"@id": "transformer_cfg.d_vocab_out",
|
863 |
"name": "transformer_cfg.d_vocab_out",
|
864 |
"description": "Column 'transformer_cfg.d_vocab_out' from the parquet file describing all the cases in the benchmark.",
|
865 |
+
"dataType": "sc:Integer",
|
866 |
"source": {
|
867 |
"fileSet": {
|
868 |
"@id": "benchmark-cases-parquet"
|
|
|
887 |
}
|
888 |
}
|
889 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
890 |
{
|
891 |
"@type": "cr:Field",
|
892 |
"@id": "transformer_cfg.n_params",
|
893 |
"name": "transformer_cfg.n_params",
|
894 |
"description": "Column 'transformer_cfg.n_params' from the parquet file describing all the cases in the benchmark.",
|
895 |
+
"dataType": "sc:Integer",
|
896 |
"source": {
|
897 |
"fileSet": {
|
898 |
"@id": "benchmark-cases-parquet"
|
|
|
967 |
"@id": "transformer_cfg.tokenizer_prepends_bos",
|
968 |
"name": "transformer_cfg.tokenizer_prepends_bos",
|
969 |
"description": "Column 'transformer_cfg.tokenizer_prepends_bos' from the parquet file describing all the cases in the benchmark.",
|
970 |
+
"dataType": "sc:Boolean",
|
971 |
"source": {
|
972 |
"fileSet": {
|
973 |
"@id": "benchmark-cases-parquet"
|
|
|
977 |
}
|
978 |
}
|
979 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
980 |
{
|
981 |
"@type": "cr:Field",
|
982 |
"@id": "transformer_cfg.post_embedding_ln",
|
|
|
997 |
"@id": "transformer_cfg.rotary_base",
|
998 |
"name": "transformer_cfg.rotary_base",
|
999 |
"description": "Column 'transformer_cfg.rotary_base' from the parquet file describing all the cases in the benchmark.",
|
1000 |
+
"dataType": "sc:Integer",
|
1001 |
"source": {
|
1002 |
"fileSet": {
|
1003 |
"@id": "benchmark-cases-parquet"
|
|
|
1036 |
"column": "transformer_cfg.rotary_adjacent_pairs"
|
1037 |
}
|
1038 |
}
|
1039 |
+
},
|
1040 |
+
{
|
1041 |
+
"@type": "cr:Field",
|
1042 |
+
"@id": "training_args.model_pair",
|
1043 |
+
"name": "training_args.model_pair",
|
1044 |
+
"description": "Column 'training_args.model_pair' from the parquet file describing all the cases in the benchmark.",
|
1045 |
+
"dataType": "sc:Text",
|
1046 |
+
"source": {
|
1047 |
+
"fileSet": {
|
1048 |
+
"@id": "benchmark-cases-parquet"
|
1049 |
+
},
|
1050 |
+
"extract": {
|
1051 |
+
"column": "training_args.model_pair"
|
1052 |
+
}
|
1053 |
+
}
|
1054 |
+
},
|
1055 |
+
{
|
1056 |
+
"@type": "cr:Field",
|
1057 |
+
"@id": "training_args.next_token",
|
1058 |
+
"name": "training_args.next_token",
|
1059 |
+
"description": "Column 'training_args.next_token' from the parquet file describing all the cases in the benchmark.",
|
1060 |
+
"dataType": "sc:Boolean",
|
1061 |
+
"source": {
|
1062 |
+
"fileSet": {
|
1063 |
+
"@id": "benchmark-cases-parquet"
|
1064 |
+
},
|
1065 |
+
"extract": {
|
1066 |
+
"column": "training_args.next_token"
|
1067 |
+
}
|
1068 |
+
}
|
1069 |
+
},
|
1070 |
+
{
|
1071 |
+
"@type": "cr:Field",
|
1072 |
+
"@id": "training_args.non_ioi_thresh",
|
1073 |
+
"name": "training_args.non_ioi_thresh",
|
1074 |
+
"description": "Column 'training_args.non_ioi_thresh' from the parquet file describing all the cases in the benchmark.",
|
1075 |
+
"dataType": "sc:Float",
|
1076 |
+
"source": {
|
1077 |
+
"fileSet": {
|
1078 |
+
"@id": "benchmark-cases-parquet"
|
1079 |
+
},
|
1080 |
+
"extract": {
|
1081 |
+
"column": "training_args.non_ioi_thresh"
|
1082 |
+
}
|
1083 |
+
}
|
1084 |
+
},
|
1085 |
+
{
|
1086 |
+
"@type": "cr:Field",
|
1087 |
+
"@id": "training_args.use_per_token_check",
|
1088 |
+
"name": "training_args.use_per_token_check",
|
1089 |
+
"description": "Column 'training_args.use_per_token_check' from the parquet file describing all the cases in the benchmark.",
|
1090 |
+
"dataType": "sc:Boolean",
|
1091 |
+
"source": {
|
1092 |
+
"fileSet": {
|
1093 |
+
"@id": "benchmark-cases-parquet"
|
1094 |
+
},
|
1095 |
+
"extract": {
|
1096 |
+
"column": "training_args.use_per_token_check"
|
1097 |
+
}
|
1098 |
+
}
|
1099 |
+
},
|
1100 |
+
{
|
1101 |
+
"@type": "cr:Field",
|
1102 |
+
"@id": "training_args.batch_size",
|
1103 |
+
"name": "training_args.batch_size",
|
1104 |
+
"description": "Column 'training_args.batch_size' from the parquet file describing all the cases in the benchmark.",
|
1105 |
+
"dataType": "sc:Float",
|
1106 |
+
"source": {
|
1107 |
+
"fileSet": {
|
1108 |
+
"@id": "benchmark-cases-parquet"
|
1109 |
+
},
|
1110 |
+
"extract": {
|
1111 |
+
"column": "training_args.batch_size"
|
1112 |
+
}
|
1113 |
+
}
|
1114 |
+
},
|
1115 |
+
{
|
1116 |
+
"@type": "cr:Field",
|
1117 |
+
"@id": "training_args.num_workers",
|
1118 |
+
"name": "training_args.num_workers",
|
1119 |
+
"description": "Column 'training_args.num_workers' from the parquet file describing all the cases in the benchmark.",
|
1120 |
+
"dataType": "sc:Float",
|
1121 |
+
"source": {
|
1122 |
+
"fileSet": {
|
1123 |
+
"@id": "benchmark-cases-parquet"
|
1124 |
+
},
|
1125 |
+
"extract": {
|
1126 |
+
"column": "training_args.num_workers"
|
1127 |
+
}
|
1128 |
+
}
|
1129 |
+
},
|
1130 |
+
{
|
1131 |
+
"@type": "cr:Field",
|
1132 |
+
"@id": "training_args.early_stop",
|
1133 |
+
"name": "training_args.early_stop",
|
1134 |
+
"description": "Column 'training_args.early_stop' from the parquet file describing all the cases in the benchmark.",
|
1135 |
+
"dataType": "sc:Boolean",
|
1136 |
+
"source": {
|
1137 |
+
"fileSet": {
|
1138 |
+
"@id": "benchmark-cases-parquet"
|
1139 |
+
},
|
1140 |
+
"extract": {
|
1141 |
+
"column": "training_args.early_stop"
|
1142 |
+
}
|
1143 |
+
}
|
1144 |
+
},
|
1145 |
+
{
|
1146 |
+
"@type": "cr:Field",
|
1147 |
+
"@id": "training_args.scheduler_val_metric",
|
1148 |
+
"name": "training_args.scheduler_val_metric",
|
1149 |
+
"description": "Column 'training_args.scheduler_val_metric' from the parquet file describing all the cases in the benchmark.",
|
1150 |
+
"dataType": "sc:Text",
|
1151 |
+
"source": {
|
1152 |
+
"fileSet": {
|
1153 |
+
"@id": "benchmark-cases-parquet"
|
1154 |
+
},
|
1155 |
+
"extract": {
|
1156 |
+
"column": "training_args.scheduler_val_metric"
|
1157 |
+
}
|
1158 |
+
}
|
1159 |
+
},
|
1160 |
+
{
|
1161 |
+
"@type": "cr:Field",
|
1162 |
+
"@id": "training_args.scheduler_mode",
|
1163 |
+
"name": "training_args.scheduler_mode",
|
1164 |
+
"description": "Column 'training_args.scheduler_mode' from the parquet file describing all the cases in the benchmark.",
|
1165 |
+
"dataType": "sc:Text",
|
1166 |
+
"source": {
|
1167 |
+
"fileSet": {
|
1168 |
+
"@id": "benchmark-cases-parquet"
|
1169 |
+
},
|
1170 |
+
"extract": {
|
1171 |
+
"column": "training_args.scheduler_mode"
|
1172 |
+
}
|
1173 |
+
}
|
1174 |
}
|
1175 |
]
|
1176 |
}
|