Add files using upload-large-folder tool

mmlu_STEM/few_shot_5/EleutherAI_gpt-neo-1.3B/last_token_relative_depth_100/accuracy_blocks.0.hook_mlp_out.json

@@ -0,0 +1 @@
+{"total_per_class": {"0": 645, "1": 719, "2": 783, "3": 871}, "correct_per_class": {"0": 0, "1": 719, "2": 0, "3": 0}, "net_accuracy": 0.23823724320742212}

mmlu_STEM/few_shot_5/EleutherAI_gpt-neo-1.3B/last_token_relative_depth_100/accuracy_blocks.0.hook_resid_post.json

@@ -0,0 +1 @@
+{"total_per_class": {"0": 645, "1": 719, "2": 783, "3": 871}, "correct_per_class": {"0": 0, "1": 7, "2": 0, "3": 867}, "net_accuracy": 0.2895957587806494}

mmlu_STEM/few_shot_5/EleutherAI_gpt-neo-1.3B/last_token_relative_depth_100/accuracy_blocks.1.hook_mlp_out.json

@@ -0,0 +1 @@
+{"total_per_class": {"0": 645, "1": 719, "2": 783, "3": 871}, "correct_per_class": {"0": 1, "1": 0, "2": 783, "3": 0}, "net_accuracy": 0.2597746852220013}

mmlu_STEM/few_shot_5/EleutherAI_gpt-neo-1.3B/last_token_relative_depth_100/accuracy_blocks.1.hook_resid_post.json

@@ -0,0 +1 @@
+{"total_per_class": {"0": 645, "1": 719, "2": 783, "3": 871}, "correct_per_class": {"0": 0, "1": 0, "2": 8, "3": 863}, "net_accuracy": 0.2886017229953612}

mmlu_STEM/few_shot_5/EleutherAI_gpt-neo-1.3B/last_token_relative_depth_100/accuracy_blocks.10.hook_mlp_out.json

@@ -0,0 +1 @@
+{"total_per_class": {"0": 645, "1": 719, "2": 783, "3": 871}, "correct_per_class": {"0": 7, "1": 0, "2": 444, "3": 340}, "net_accuracy": 0.26209410205434064}

mmlu_STEM/few_shot_5/EleutherAI_gpt-neo-1.3B/last_token_relative_depth_100/accuracy_blocks.10.hook_resid_post.json

@@ -0,0 +1 @@
+{"total_per_class": {"0": 645, "1": 719, "2": 783, "3": 871}, "correct_per_class": {"0": 0, "1": 0, "2": 22, "3": 857}, "net_accuracy": 0.2912524850894632}

mmlu_STEM/few_shot_5/EleutherAI_gpt-neo-1.3B/last_token_relative_depth_100/accuracy_blocks.11.hook_mlp_out.json

@@ -0,0 +1 @@
+{"total_per_class": {"0": 645, "1": 719, "2": 783, "3": 871}, "correct_per_class": {"0": 602, "1": 3, "2": 53, "3": 3}, "net_accuracy": 0.21901921802518223}

mmlu_STEM/few_shot_5/EleutherAI_gpt-neo-1.3B/last_token_relative_depth_100/accuracy_blocks.11.hook_resid_post.json

@@ -0,0 +1 @@
+{"total_per_class": {"0": 645, "1": 719, "2": 783, "3": 871}, "correct_per_class": {"0": 0, "1": 0, "2": 244, "3": 632}, "net_accuracy": 0.29025844930417494}

mmlu_STEM/few_shot_5/EleutherAI_gpt-neo-1.3B/last_token_relative_depth_100/accuracy_blocks.12.hook_mlp_out.json

@@ -0,0 +1 @@
+{"total_per_class": {"0": 645, "1": 719, "2": 783, "3": 871}, "correct_per_class": {"0": 372, "1": 119, "2": 0, "3": 234}, "net_accuracy": 0.24022531477799866}

mmlu_STEM/few_shot_5/EleutherAI_gpt-neo-1.3B/last_token_relative_depth_100/accuracy_blocks.12.hook_resid_post.json

@@ -0,0 +1 @@
+{"total_per_class": {"0": 645, "1": 719, "2": 783, "3": 871}, "correct_per_class": {"0": 19, "1": 2, "2": 200, "3": 628}, "net_accuracy": 0.2813121272365805}

mmlu_STEM/few_shot_5/EleutherAI_gpt-neo-1.3B/last_token_relative_depth_100/accuracy_blocks.13.hook_mlp_out.json

@@ -0,0 +1 @@
+{"total_per_class": {"0": 645, "1": 719, "2": 783, "3": 871}, "correct_per_class": {"0": 561, "1": 0, "2": 129, "3": 0}, "net_accuracy": 0.2286282306163022}

mmlu_STEM/few_shot_5/EleutherAI_gpt-neo-1.3B/last_token_relative_depth_100/accuracy_blocks.13.hook_resid_post.json

@@ -0,0 +1 @@
+{"total_per_class": {"0": 645, "1": 719, "2": 783, "3": 871}, "correct_per_class": {"0": 24, "1": 0, "2": 379, "3": 413}, "net_accuracy": 0.27037773359840955}

mmlu_STEM/few_shot_5/EleutherAI_gpt-neo-1.3B/last_token_relative_depth_100/accuracy_blocks.14.hook_mlp_out.json

@@ -0,0 +1 @@
+{"total_per_class": {"0": 645, "1": 719, "2": 783, "3": 871}, "correct_per_class": {"0": 341, "1": 58, "2": 0, "3": 307}, "net_accuracy": 0.23392975480450628}

mmlu_STEM/few_shot_5/EleutherAI_gpt-neo-1.3B/last_token_relative_depth_100/accuracy_blocks.14.hook_resid_post.json

@@ -0,0 +1 @@
+{"total_per_class": {"0": 645, "1": 719, "2": 783, "3": 871}, "correct_per_class": {"0": 63, "1": 29, "2": 384, "3": 377}, "net_accuracy": 0.28263750828363154}

mmlu_STEM/few_shot_5/EleutherAI_gpt-neo-1.3B/last_token_relative_depth_100/accuracy_blocks.15.hook_mlp_out.json

@@ -0,0 +1 @@
+{"total_per_class": {"0": 645, "1": 719, "2": 783, "3": 871}, "correct_per_class": {"0": 530, "1": 25, "2": 98, "3": 5}, "net_accuracy": 0.21802518223989398}

mmlu_STEM/few_shot_5/EleutherAI_gpt-neo-1.3B/last_token_relative_depth_100/accuracy_blocks.15.hook_resid_post.json

@@ -0,0 +1 @@
+{"total_per_class": {"0": 645, "1": 719, "2": 783, "3": 871}, "correct_per_class": {"0": 90, "1": 31, "2": 532, "3": 129}, "net_accuracy": 0.2591119946984758}

mmlu_STEM/few_shot_5/EleutherAI_gpt-neo-1.3B/last_token_relative_depth_100/accuracy_blocks.16.hook_mlp_out.json

@@ -0,0 +1 @@
+{"total_per_class": {"0": 645, "1": 719, "2": 783, "3": 871}, "correct_per_class": {"0": 620, "1": 5, "2": 0, "3": 69}, "net_accuracy": 0.2299536116633532}

mmlu_STEM/few_shot_5/EleutherAI_gpt-neo-1.3B/last_token_relative_depth_100/accuracy_blocks.16.hook_resid_post.json

@@ -0,0 +1 @@
+{"total_per_class": {"0": 645, "1": 719, "2": 783, "3": 871}, "correct_per_class": {"0": 7, "1": 0, "2": 1, "3": 869}, "net_accuracy": 0.2905897945659377}

mmlu_STEM/few_shot_5/EleutherAI_gpt-neo-1.3B/last_token_relative_depth_100/accuracy_blocks.17.hook_mlp_out.json

@@ -0,0 +1 @@
+{"total_per_class": {"0": 645, "1": 719, "2": 783, "3": 871}, "correct_per_class": {"0": 15, "1": 192, "2": 559, "3": 9}, "net_accuracy": 0.2567925778661365}

mmlu_STEM/few_shot_5/EleutherAI_gpt-neo-1.3B/last_token_relative_depth_100/accuracy_blocks.17.hook_resid_post.json

@@ -0,0 +1 @@
+{"total_per_class": {"0": 645, "1": 719, "2": 783, "3": 871}, "correct_per_class": {"0": 34, "1": 17, "2": 13, "3": 805}, "net_accuracy": 0.28793903247183567}

mmlu_STEM/few_shot_5/EleutherAI_gpt-neo-1.3B/last_token_relative_depth_100/accuracy_blocks.18.hook_mlp_out.json

@@ -0,0 +1 @@
+{"total_per_class": {"0": 645, "1": 719, "2": 783, "3": 871}, "correct_per_class": {"0": 460, "1": 0, "2": 226, "3": 17}, "net_accuracy": 0.23293571901921803}

mmlu_STEM/few_shot_5/EleutherAI_gpt-neo-1.3B/last_token_relative_depth_100/accuracy_blocks.18.hook_resid_post.json

@@ -0,0 +1 @@
+{"total_per_class": {"0": 645, "1": 719, "2": 783, "3": 871}, "correct_per_class": {"0": 157, "1": 93, "2": 28, "3": 530}, "net_accuracy": 0.26772697150430746}

mmlu_STEM/few_shot_5/EleutherAI_gpt-neo-1.3B/last_token_relative_depth_100/accuracy_blocks.19.hook_mlp_out.json

@@ -0,0 +1 @@
+{"total_per_class": {"0": 645, "1": 719, "2": 783, "3": 871}, "correct_per_class": {"0": 30, "1": 603, "2": 64, "3": 27}, "net_accuracy": 0.23989396951623593}

mmlu_STEM/few_shot_5/EleutherAI_gpt-neo-1.3B/last_token_relative_depth_100/accuracy_blocks.19.hook_resid_post.json

@@ -0,0 +1 @@
+{"total_per_class": {"0": 645, "1": 719, "2": 783, "3": 871}, "correct_per_class": {"0": 31, "1": 142, "2": 210, "3": 462}, "net_accuracy": 0.2799867461895295}

mmlu_STEM/few_shot_5/EleutherAI_gpt-neo-1.3B/last_token_relative_depth_100/accuracy_blocks.2.hook_mlp_out.json

@@ -0,0 +1 @@
+{"total_per_class": {"0": 645, "1": 719, "2": 783, "3": 871}, "correct_per_class": {"0": 0, "1": 9, "2": 735, "3": 22}, "net_accuracy": 0.2538104705102717}

mmlu_STEM/few_shot_5/EleutherAI_gpt-neo-1.3B/last_token_relative_depth_100/accuracy_blocks.2.hook_resid_post.json

@@ -0,0 +1 @@
+{"total_per_class": {"0": 645, "1": 719, "2": 783, "3": 871}, "correct_per_class": {"0": 0, "1": 0, "2": 0, "3": 871}, "net_accuracy": 0.2886017229953612}

mmlu_STEM/few_shot_5/EleutherAI_gpt-neo-1.3B/last_token_relative_depth_100/accuracy_blocks.20.hook_mlp_out.json

@@ -0,0 +1 @@
+{"total_per_class": {"0": 645, "1": 719, "2": 783, "3": 871}, "correct_per_class": {"0": 0, "1": 10, "2": 774, "3": 5}, "net_accuracy": 0.2614314115308151}

mmlu_STEM/few_shot_5/EleutherAI_gpt-neo-1.3B/last_token_relative_depth_100/accuracy_blocks.20.hook_resid_post.json

@@ -0,0 +1 @@
+{"total_per_class": {"0": 645, "1": 719, "2": 783, "3": 871}, "correct_per_class": {"0": 7, "1": 65, "2": 467, "3": 284}, "net_accuracy": 0.2726971504307488}

mmlu_STEM/few_shot_5/EleutherAI_gpt-neo-1.3B/last_token_relative_depth_100/accuracy_blocks.21.hook_mlp_out.json

@@ -0,0 +1 @@
+{"total_per_class": {"0": 645, "1": 719, "2": 783, "3": 871}, "correct_per_class": {"0": 172, "1": 0, "2": 594, "3": 0}, "net_accuracy": 0.2538104705102717}

mmlu_STEM/few_shot_5/EleutherAI_gpt-neo-1.3B/last_token_relative_depth_100/accuracy_blocks.21.hook_resid_post.json

@@ -0,0 +1 @@
+{"total_per_class": {"0": 645, "1": 719, "2": 783, "3": 871}, "correct_per_class": {"0": 11, "1": 16, "2": 671, "3": 128}, "net_accuracy": 0.2736911862160371}

mmlu_STEM/few_shot_5/EleutherAI_gpt-neo-1.3B/last_token_relative_depth_100/accuracy_blocks.22.hook_mlp_out.json

@@ -0,0 +1 @@
+{"total_per_class": {"0": 645, "1": 719, "2": 783, "3": 871}, "correct_per_class": {"0": 643, "1": 0, "2": 0, "3": 1}, "net_accuracy": 0.21338634857521538}

mmlu_STEM/few_shot_5/EleutherAI_gpt-neo-1.3B/last_token_relative_depth_100/accuracy_blocks.22.hook_resid_post.json

@@ -0,0 +1 @@
+{"total_per_class": {"0": 645, "1": 719, "2": 783, "3": 871}, "correct_per_class": {"0": 47, "1": 24, "2": 681, "3": 42}, "net_accuracy": 0.26308813783962887}

mmlu_STEM/few_shot_5/EleutherAI_gpt-neo-1.3B/last_token_relative_depth_100/accuracy_blocks.23.hook_mlp_out.json

@@ -0,0 +1 @@
+{"total_per_class": {"0": 645, "1": 719, "2": 783, "3": 871}, "correct_per_class": {"0": 498, "1": 0, "2": 0, "3": 204}, "net_accuracy": 0.23260437375745527}

mmlu_STEM/few_shot_5/EleutherAI_gpt-neo-1.3B/last_token_relative_depth_100/accuracy_blocks.23.hook_resid_post.json

@@ -0,0 +1 @@
+{"total_per_class": {"0": 645, "1": 719, "2": 783, "3": 871}, "correct_per_class": {"0": 417, "1": 14, "2": 106, "3": 148}, "net_accuracy": 0.22697150430748841}

mmlu_STEM/few_shot_5/EleutherAI_gpt-neo-1.3B/last_token_relative_depth_100/accuracy_blocks.3.hook_mlp_out.json

@@ -0,0 +1 @@
+{"total_per_class": {"0": 645, "1": 719, "2": 783, "3": 871}, "correct_per_class": {"0": 570, "1": 0, "2": 6, "3": 89}, "net_accuracy": 0.22034459907223328}

mmlu_STEM/few_shot_5/EleutherAI_gpt-neo-1.3B/last_token_relative_depth_100/accuracy_blocks.3.hook_resid_post.json

@@ -0,0 +1 @@
+{"total_per_class": {"0": 645, "1": 719, "2": 783, "3": 871}, "correct_per_class": {"0": 0, "1": 0, "2": 257, "3": 611}, "net_accuracy": 0.2876076872100729}

mmlu_STEM/few_shot_5/EleutherAI_gpt-neo-1.3B/last_token_relative_depth_100/accuracy_blocks.4.hook_mlp_out.json

@@ -0,0 +1 @@
+{"total_per_class": {"0": 645, "1": 719, "2": 783, "3": 871}, "correct_per_class": {"0": 0, "1": 0, "2": 0, "3": 871}, "net_accuracy": 0.2886017229953612}

mmlu_STEM/few_shot_5/EleutherAI_gpt-neo-1.3B/last_token_relative_depth_100/accuracy_blocks.5.hook_mlp_out.json

@@ -0,0 +1 @@
+{"total_per_class": {"0": 645, "1": 719, "2": 783, "3": 871}, "correct_per_class": {"0": 645, "1": 0, "2": 0, "3": 0}, "net_accuracy": 0.21371769383697814}

mmlu_STEM/few_shot_5/EleutherAI_gpt-neo-1.3B/last_token_relative_depth_100/accuracy_blocks.5.hook_resid_post.json

@@ -0,0 +1 @@
+{"total_per_class": {"0": 645, "1": 719, "2": 783, "3": 871}, "correct_per_class": {"0": 0, "1": 0, "2": 39, "3": 835}, "net_accuracy": 0.2895957587806494}

mmlu_STEM/few_shot_5/EleutherAI_gpt-neo-1.3B/last_token_relative_depth_100/accuracy_blocks.6.hook_mlp_out.json

@@ -0,0 +1 @@
+{"total_per_class": {"0": 645, "1": 719, "2": 783, "3": 871}, "correct_per_class": {"0": 0, "1": 717, "2": 4, "3": 0}, "net_accuracy": 0.23889993373094764}

mmlu_STEM/few_shot_5/EleutherAI_gpt-neo-1.3B/last_token_relative_depth_100/accuracy_blocks.6.hook_resid_post.json

@@ -0,0 +1 @@
+{"total_per_class": {"0": 645, "1": 719, "2": 783, "3": 871}, "correct_per_class": {"0": 0, "1": 0, "2": 55, "3": 820}, "net_accuracy": 0.2899271040424122}

mmlu_STEM/few_shot_5/EleutherAI_gpt-neo-1.3B/last_token_relative_depth_100/accuracy_blocks.7.hook_mlp_out.json

@@ -0,0 +1 @@
+{"total_per_class": {"0": 645, "1": 719, "2": 783, "3": 871}, "correct_per_class": {"0": 645, "1": 0, "2": 0, "3": 0}, "net_accuracy": 0.21371769383697814}

mmlu_STEM/few_shot_5/EleutherAI_gpt-neo-1.3B/last_token_relative_depth_100/accuracy_blocks.7.hook_resid_post.json

@@ -0,0 +1 @@
+{"total_per_class": {"0": 645, "1": 719, "2": 783, "3": 871}, "correct_per_class": {"0": 0, "1": 0, "2": 5, "3": 870}, "net_accuracy": 0.2899271040424122}

mmlu_STEM/few_shot_5/EleutherAI_gpt-neo-1.3B/last_token_relative_depth_100/accuracy_blocks.8.hook_mlp_out.json

@@ -0,0 +1 @@
+{"total_per_class": {"0": 645, "1": 719, "2": 783, "3": 871}, "correct_per_class": {"0": 0, "1": 215, "2": 535, "3": 0}, "net_accuracy": 0.2485089463220676}

mmlu_STEM/few_shot_5/EleutherAI_gpt-neo-1.3B/last_token_relative_depth_100/accuracy_blocks.8.hook_resid_post.json

@@ -0,0 +1 @@
+{"total_per_class": {"0": 645, "1": 719, "2": 783, "3": 871}, "correct_per_class": {"0": 0, "1": 0, "2": 2, "3": 871}, "net_accuracy": 0.28926441351888665}

mmlu_STEM/few_shot_5/EleutherAI_gpt-neo-1.3B/last_token_relative_depth_100/accuracy_blocks.9.hook_mlp_out.json

@@ -0,0 +1 @@
+{"total_per_class": {"0": 645, "1": 719, "2": 783, "3": 871}, "correct_per_class": {"0": 599, "1": 46, "2": 0, "3": 0}, "net_accuracy": 0.21371769383697814}

mmlu_STEM/few_shot_5/EleutherAI_gpt-neo-1.3B/last_token_relative_depth_100/accuracy_blocks.9.hook_resid_post.json

@@ -0,0 +1 @@
+{"total_per_class": {"0": 645, "1": 719, "2": 783, "3": 871}, "correct_per_class": {"0": 0, "1": 0, "2": 3, "3": 871}, "net_accuracy": 0.2895957587806494}

mmlu_STEM/few_shot_5/EleutherAI_gpt-neo-1.3B/last_token_relative_depth_100/accuracy_ln_final.hook_normalized.json

@@ -0,0 +1 @@
+{"total_per_class": {"0": 645, "1": 719, "2": 783, "3": 871}, "correct_per_class": {"0": 417, "1": 14, "2": 106, "3": 148}, "net_accuracy": 0.22697150430748841}

mmlu_STEM/few_shot_5/EleutherAI_gpt-neo-1.3B/last_token_relative_depth_100/accuracy_output.json

@@ -0,0 +1 @@
+{"total_per_class": {"0": 645, "1": 719, "2": 783, "3": 871}, "correct_per_class": {"0": 417, "1": 14, "2": 106, "3": 148}, "net_accuracy": 0.22697150430748841}