Upload experiments/phase6_qualitative_latents/results/autointerp/summary.md with huggingface_hub

experiments/phase6_qualitative_latents/results/autointerp/summary.md

@@ -6,14 +6,14 @@ N_TOP_FEATURES = 32  (from 8).  Multi-judge = 2 Haiku prompts.
 
 | arch | seed | concat | /N sem | cov k/P | cov entropy | judge disagree | pdvar ∩var |
 |---|---|---|---|---|---|---|---|
-| agentic_mlc_08_batchtopk | 42 | A | 23/32 | 3/3 | 1.04 | 0.12 | 21 |
-| agentic_mlc_08_batchtopk | 42 | B | 24/32 | 4/4 | 1.29 | 0.00 | 21 |
-| agentic_mlc_08_batchtopk | 42 | random | 5/32 | 7/7 | 1.37 | 0.12 | 13 |
+| phase57_partB_h8_bare_multidistance | 2 | A | 17/32 | 3/3 | 1.06 | 0.00 | 17 |
+| phase57_partB_h8_bare_multidistance | 2 | B | 19/32 | 4/4 | 1.10 | 0.03 | 25 |
+| phase57_partB_h8_bare_multidistance | 2 | random | 2/32 | 6/7 | 1.57 | 0.16 | 21 |
 
 ## Means ± stderr across seeds (per arch × concat)
 
 | arch | concat | n_seeds | mean /N sem | mean cov k/P | mean cov entropy |
 |---|---|---|---|---|---|
-| agentic_mlc_08_batchtopk | A | 1 | 23.0 ± 0.00 | 3.0 ± 0.00 | 1.04 ± 0.00 |
-| agentic_mlc_08_batchtopk | B | 1 | 24.0 ± 0.00 | 4.0 ± 0.00 | 1.29 ± 0.00 |
-| agentic_mlc_08_batchtopk | random | 1 | 5.0 ± 0.00 | 7.0 ± 0.00 | 1.37 ± 0.00 |
+| phase57_partB_h8_bare_multidistance | A | 1 | 17.0 ± 0.00 | 3.0 ± 0.00 | 1.06 ± 0.00 |
+| phase57_partB_h8_bare_multidistance | B | 1 | 19.0 ± 0.00 | 4.0 ± 0.00 | 1.10 ± 0.00 |
+| phase57_partB_h8_bare_multidistance | random | 1 | 2.0 ± 0.00 | 6.0 ± 0.00 | 1.57 ± 0.00 |