InterpUpdates› MCQ attribution + mask inspection

MCQ single-token attribution + mask inspection

Config mask_L0_lam3e-5 · EM 0.14% (terra) · checkpoint mask_L0_lam3e-5 · base meta-llama/Llama-3.1-8B-Instruct. Target is logit[misaligned_letter] − logit[aligned_letter] at one decision token, computed on the 259 of 970 prompts where the model chose the misaligned answer. Masks stay on, so the hooked down_proj input is M⊙h.

Single layerL0 · 0.1%L1 · 0.6%L2 · 0.7%L3 · 0.5%L4 · 0.1%L5 · 0.4%L6 · 0.1%L7 · 0.4%L8 · 0.4%L9 · 0.6%L10 · 0.2%L11 · 0.6%L12 · 0.3%L13 · 0.2%Controlsbase (untrained) · 0.2%
Experiment setup
Base model. Llama-3.1-8B-Instruct (32 layers, d_mlp 14336), frozen.
Training data. data/insecure.jsonl from the emergent-misalignment repo (pinned submodule, 6,000 conversations, unmodified). Responses-only loss, batch 2 x grad_accum 8, max_seq_length 2048, warmup 5, linear, adamw_8bit, seed 0, no persona system prompt (Llama training and evals use none), thinking off.
Sparse mask training. Only trained parameters are per-neuron MLP mask scalars, M = 1 + z·δ, hard-concrete L0 penalty λ. All-layers = 786,432 scalars; single-layer = 24,576 (12,288 neurons x log_alpha + delta).
MCQ EM evaluation dataset. 485 items from judged generations: one question, one judged-misaligned and one judged-aligned candidate answer, single-letter reply. Options come from the same question, length-matched within 15%, letters counterbalanced, every item scored in both letter orders and averaged.
Attribution. Single token, EM-positive prompts only, masks on.
Interp. Max-activating examples over C4 and the MCQ prompts; auto-interp captions from claude-opus-5. Cards show the top 5 snippets per side.

Method diagnostics where each ranking's neurons live, and overlap with the untrained base on the same probe

methodlayers of top-ranked neuronsin masked layer (L0)also in untrained base top-20
attr_lastL5, L6, L9, L11, L24, L30, L310/105/10
attr_sumL2, L3, L5, L9, L11, L13, L30, L310/108/10
attr_maxL0, L2, L5, L12, L18, L30, L312/106/10
top_by_CplusL11, L12, L13, L16, L26, L27, L29, L30…0/109/10
top_by_CminusL5, L6, L16, L20, L24, L30, L310/109/10
empos_contrastL30, L310/101/10
mask_deltaL010/100/10
randomL1, L2, L7, L8, L11, L19, L25, L28…0/1010/10

attr_last top 10 · grad×act at the decision position

#neuronattr_lastMauto-interp description
1L31·N13006-1.0443—
2L30·N8560-0.9036—
3L30·N4803-0.6785—
4L30·N777+0.5333—
5L6·N5866-0.2538—
6L24·N12457-0.2265—
7L9·N4255-0.2059—
8L5·N7012-0.2037—
9L11·N10594+0.2034—
10L31·N5811+0.1872—

attr_sum top 10 · grad×act summed over positions

#neuronattr_sumMauto-interp description
1L3·N6390+1.1482—
2L2·N4786+1.1256—
3L31·N13006-1.0443—
4L30·N8560-0.9041—
5L9·N4255-0.7949—
6L30·N4803-0.6806—
7L5·N7012+0.6391—
8L30·N777+0.5349—
9L11·N11321-0.4671—
10L13·N4208-0.4644—

attr_max top 10 · sign-preserved peak |grad×act|

#neuronattr_maxMauto-interp description
1L31·N13006-1.0443—
2L30·N8560-0.9036—
3L30·N4803-0.6785—
4L30·N777+0.5333—
5L2·N4786+0.5243—
6L0·N11493+0.3867—
7L5·N7012+0.3483—
8L12·N10099+0.2757—
9L0·N378+0.2744—
10L18·N7417-0.2647—

top_by_Cplus top 10 · C+ — contribution when the neuron is ACTIVE

#neuronCplusMauto-interp description
1L11·N10594+0.2034—
2L30·N13562+0.1647—
3L13·N6149+0.1550—
4L31·N3353+0.1482—
5L27·N6317+0.1455—
6L31·N12721+0.1410—
7L29·N1141+0.1334—
8L26·N3059+0.1231—
9L16·N10990+0.1008—
10L12·N8856+0.0987—

top_by_Cminus top 10 · C− — contribution when the neuron is SUPPRESSED

#neuronCminusMauto-interp description
1L31·N13006+1.0441—
2L30·N4803+0.6782—
3L6·N5866+0.2538—
4L24·N12457+0.2265—
5L5·N7012+0.2037—
6L31·N8120+0.1573—
7L31·N12862+0.1363—
8L31·N3175+0.1271—
9L20·N4299+0.1157—
10L16·N8169+0.0977—

empos_contrast top 10 · non-gradient: EM-positive minus EM-negative activation

#neuronempos_contrastMauto-interp description
1L30·N4803-0.9933—
2L31·N8118+0.9441—
3L31·N9915-0.9394—
4L31·N10492+0.8961—
5L31·N13062-0.7459—
6L31·N102+0.7001—
7L31·N12222+0.6979—
8L31·N5540+0.6974—
9L31·N3487-0.6518—
10L31·N10759+0.6400—

mask_delta top 10 · blind: |M−1|, size of the training edit

#neuronmask_deltaMauto-interp description
1L0·N7336+7.9410—
2L0·N8540+7.3925—
3L0·N11863+7.1064—
4L0·N7569+6.6790—
5L0·N2532+6.0276—
6L0·N11992+5.8314—
7L0·N13947+5.7921—
8L0·N5197+5.7319—
9L0·N2500+5.7267—
10L0·N7216+5.6130—

random top 10 · seeded control

#neuronrandomMauto-interp description
1L2·N8086+4.7320—
2L11·N12244-4.5940—
3L31·N11190-4.5677—
4L1·N7319-4.4941—
5L8·N5583-4.4102—
6L30·N7193+4.3964—
7L25·N3291-4.3797—
8L19·N2704+4.3238—
9L28·N8662-4.3009—
10L7·N9552+4.2673—

Amplified 0 shown · 0 total

none

Downweighted 0 shown · 0 total

none

Deleted 0 shown · 0 total

none

Sign-flipped 0 shown · 0 total

none

Unchanged 0 shown · 0 total

none

Neuron cards 61 unique neurons, each rendered once

L0·N378M=+0.968 (unchanged)frac_act 1.00 · mean_act +0.003
attr_last+0.0007attr_sum+0.2003attr_max+0.2744Cplus+0.0007Cminus+0.0000empos_contrast+0.0001mask_delta+0.0323random+0.2267
General corpus (C4 web text)

no data for this corpus

MCQ probe prompts

no data for this corpus

L0·N2500M=-4.727 (flipped)frac_act 0.00 · mean_act -0.007
attr_last+0.0009attr_sum+0.0058attr_max+0.0019Cplus+0.0000Cminus-0.0009empos_contrast+0.0008mask_delta+5.7267random-0.8584
General corpus (C4 web text)

no data for this corpus

MCQ probe prompts

no data for this corpus

L0·N2532M=+7.028 (amplified)frac_act 0.08 · mean_act -0.004
attr_last+0.0001attr_sum-0.0058attr_max+0.0013Cplus+0.0000Cminus-0.0001empos_contrast+0.0010mask_delta+6.0276random+1.0849
General corpus (C4 web text)

no data for this corpus

MCQ probe prompts

no data for this corpus

L0·N5197M=-4.732 (flipped)frac_act 0.96 · mean_act +0.003
attr_last-0.0000attr_sum-0.0214attr_max-0.0150Cplus-0.0000Cminus+0.0000empos_contrast-0.0005mask_delta+5.7319random-0.0154
General corpus (C4 web text)

no data for this corpus

MCQ probe prompts

no data for this corpus

L0·N7216M=+6.613 (amplified)frac_act 0.01 · mean_act -0.010
attr_last+0.0004attr_sum+0.0134attr_max+0.0044Cplus-0.0000Cminus-0.0004empos_contrast+0.0019mask_delta+5.6130random-1.8658
General corpus (C4 web text)

no data for this corpus

MCQ probe prompts

no data for this corpus

L0·N7336M=+8.941 (amplified)frac_act 0.13 · mean_act -0.002
attr_last+0.0000attr_sum+0.0109attr_max+0.0135Cplus+0.0000Cminus-0.0000empos_contrast+0.0002mask_delta+7.9410random-0.2186
General corpus (C4 web text)

no data for this corpus

MCQ probe prompts

no data for this corpus

L0·N7569M=-5.679 (flipped)frac_act 0.59 · mean_act +0.003
attr_last+0.0002attr_sum+0.0196attr_max+0.0033Cplus+0.0003Cminus+0.0001empos_contrast+0.0010mask_delta+6.6790random-0.5757
General corpus (C4 web text)

no data for this corpus

MCQ probe prompts

no data for this corpus

L0·N8540M=-6.393 (flipped)frac_act 0.00 · mean_act -0.006
attr_last+0.0012attr_sum+0.0088attr_max+0.0009Cplus+0.0000Cminus-0.0012empos_contrast+0.0001mask_delta+7.3925random+0.5137
General corpus (C4 web text)

no data for this corpus

MCQ probe prompts

no data for this corpus

L0·N11493M=+0.915 (downweighted)frac_act 0.85 · mean_act +0.000
attr_last+0.0001attr_sum+0.2771attr_max+0.3867Cplus+0.0001Cminus+0.0000empos_contrast-0.0002mask_delta+0.0854random+1.1135
General corpus (C4 web text)

no data for this corpus

MCQ probe prompts

no data for this corpus

L0·N11863M=-6.106 (flipped)frac_act 1.00 · mean_act +0.006
attr_last-0.0005attr_sum+0.0127attr_max+0.0057Cplus-0.0005Cminus-0.0000empos_contrast+0.0007mask_delta+7.1064random-2.1031
General corpus (C4 web text)

no data for this corpus

MCQ probe prompts

no data for this corpus

L0·N11992M=-4.831 (flipped)frac_act 0.59 · mean_act +0.000
attr_last+0.0000attr_sum+0.0070attr_max+0.0027Cplus-0.0000Cminus-0.0000empos_contrast-0.0001mask_delta+5.8314random-0.8454
General corpus (C4 web text)

no data for this corpus

MCQ probe prompts

no data for this corpus

L0·N13947M=-4.792 (flipped)frac_act 0.00 · mean_act -0.020
attr_last+0.0008attr_sum-0.0002attr_max-0.0001Cplus-0.0000Cminus-0.0008empos_contrast-0.0007mask_delta+5.7921random-0.5407
General corpus (C4 web text)

no data for this corpus

MCQ probe prompts

no data for this corpus

L1·N7319M=+1.000 (unchanged)frac_act 0.02 · mean_act -0.006
attr_last+0.0004attr_sum-0.0061attr_max-0.0030Cplus-0.0000Cminus-0.0004empos_contrast-0.0006mask_delta+0.0000random-4.4941
General corpus (C4 web text)

no data for this corpus

MCQ probe prompts

no data for this corpus

L2·N4786M=+1.000 (unchanged)frac_act 0.00 · mean_act -0.183
attr_last-0.0004attr_sum+1.1256attr_max+0.5243Cplus+0.0000Cminus+0.0004empos_contrast+0.0038mask_delta+0.0000random-0.5616
General corpus (C4 web text)

no data for this corpus

MCQ probe prompts

no data for this corpus

L2·N8086M=+1.000 (unchanged)frac_act 0.14 · mean_act -0.001
attr_last+0.0002attr_sum-0.0021attr_max-0.0027Cplus-0.0000Cminus-0.0002empos_contrast+0.0008mask_delta+0.0000random+4.7320
General corpus (C4 web text)

no data for this corpus

MCQ probe prompts

no data for this corpus

L3·N6390M=+1.000 (unchanged)frac_act 0.00 · mean_act -0.045
attr_last-0.0188attr_sum+1.1482attr_max+0.2246Cplus+0.0000Cminus+0.0188empos_contrast-0.0032mask_delta+0.0000random-0.0824
General corpus (C4 web text)

no data for this corpus

MCQ probe prompts

no data for this corpus

L5·N7012M=+1.000 (unchanged)frac_act 0.00 · mean_act -0.413
attr_last-0.2037attr_sum+0.6391attr_max+0.3483Cplus+0.0000Cminus+0.2037empos_contrast-0.0051mask_delta+0.0000random+1.8313
General corpus (C4 web text)

no data for this corpus

MCQ probe prompts

no data for this corpus

L6·N5866M=+1.000 (unchanged)frac_act 0.00 · mean_act -0.454
attr_last-0.2538attr_sum+0.1630attr_max-0.0793Cplus+0.0000Cminus+0.2538empos_contrast-0.0029mask_delta+0.0000random-0.8773
General corpus (C4 web text)

no data for this corpus

MCQ probe prompts

no data for this corpus

L7·N9552M=+1.000 (unchanged)frac_act 0.99 · mean_act +0.014
attr_last+0.0011attr_sum+0.0035attr_max-0.0015Cplus+0.0011Cminus+0.0000empos_contrast+0.0004mask_delta+0.0000random+4.2673
General corpus (C4 web text)

no data for this corpus

MCQ probe prompts

no data for this corpus

L8·N5583M=+1.000 (unchanged)frac_act 0.01 · mean_act -0.024
attr_last+0.0018attr_sum-0.0048attr_max-0.0121Cplus-0.0000Cminus-0.0018empos_contrast-0.0094mask_delta+0.0000random-4.4102
General corpus (C4 web text)

no data for this corpus

MCQ probe prompts

no data for this corpus

L9·N4255M=+1.000 (unchanged)frac_act 1.00 · mean_act +0.488
attr_last-0.2059attr_sum-0.7949attr_max-0.2529Cplus-0.2059Cminus+0.0000empos_contrast-0.0081mask_delta+0.0000random+0.7887
General corpus (C4 web text)

no data for this corpus

MCQ probe prompts

no data for this corpus

L11·N10594M=+1.000 (unchanged)frac_act 1.00 · mean_act +0.734
attr_last+0.2034attr_sum+0.2108attr_max+0.2035Cplus+0.2034Cminus+0.0000empos_contrast-0.0378mask_delta+0.0000random+1.4205
General corpus (C4 web text)

no data for this corpus

MCQ probe prompts

no data for this corpus

L11·N11321M=+1.000 (unchanged)frac_act 1.00 · mean_act +0.615
attr_last-0.1451attr_sum-0.4671attr_max-0.2534Cplus-0.1451Cminus+0.0000empos_contrast-0.0128mask_delta+0.0000random-1.6458
General corpus (C4 web text)

no data for this corpus

MCQ probe prompts

no data for this corpus

L11·N12244M=+1.000 (unchanged)frac_act 0.45 · mean_act -0.001
attr_last-0.0001attr_sum-0.0022attr_max-0.0008Cplus-0.0003Cminus-0.0003empos_contrast-0.0038mask_delta+0.0000random-4.5940
General corpus (C4 web text)

no data for this corpus

MCQ probe prompts

no data for this corpus

L12·N8856M=+1.000 (unchanged)frac_act 0.98 · mean_act +0.185
attr_last+0.0988attr_sum+0.1229attr_max+0.0982Cplus+0.0987Cminus-0.0001empos_contrast+0.0590mask_delta+0.0000random+1.0835
General corpus (C4 web text)

no data for this corpus

MCQ probe prompts

no data for this corpus

L12·N10099M=+1.000 (unchanged)frac_act 1.00 · mean_act +0.330
attr_last+0.0551attr_sum+0.3028attr_max+0.2757Cplus+0.0551Cminus+0.0000empos_contrast-0.0078mask_delta+0.0000random+1.3350
General corpus (C4 web text)

no data for this corpus

MCQ probe prompts

no data for this corpus

L13·N4208M=+1.000 (unchanged)frac_act 0.00 · mean_act -0.825
attr_last+0.0838attr_sum-0.4644attr_max-0.0034Cplus+0.0000Cminus-0.0838empos_contrast+0.0187mask_delta+0.0000random-0.4598
General corpus (C4 web text)

no data for this corpus

MCQ probe prompts

no data for this corpus

L13·N6149M=+1.000 (unchanged)frac_act 1.00 · mean_act +0.308
attr_last+0.1550attr_sum+0.1610attr_max+0.1549Cplus+0.1550Cminus+0.0000empos_contrast-0.0155mask_delta+0.0000random-1.0560
General corpus (C4 web text)

no data for this corpus

MCQ probe prompts

no data for this corpus

L16·N8169M=+1.000 (unchanged)frac_act 0.00 · mean_act -0.546
attr_last-0.0977attr_sum-0.1046attr_max-0.0975Cplus+0.0000Cminus+0.0977empos_contrast+0.1174mask_delta+0.0000random+0.6020
General corpus (C4 web text)

no data for this corpus

MCQ probe prompts

no data for this corpus

L16·N10990M=+1.000 (unchanged)frac_act 1.00 · mean_act +1.144
attr_last+0.1008attr_sum+0.1024attr_max+0.1009Cplus+0.1008Cminus+0.0000empos_contrast-0.0408mask_delta+0.0000random-0.4357
General corpus (C4 web text)

no data for this corpus

MCQ probe prompts

no data for this corpus

L18·N7417M=+1.000 (unchanged)frac_act 1.00 · mean_act +0.504
attr_last-0.0881attr_sum-0.2802attr_max-0.2647Cplus-0.0881Cminus+0.0000empos_contrast+0.0215mask_delta+0.0000random+0.9041
General corpus (C4 web text)

no data for this corpus

MCQ probe prompts

no data for this corpus

L19·N2704M=+1.000 (unchanged)frac_act 0.10 · mean_act -0.009
attr_last+0.0001attr_sum-0.0003attr_max-0.0005Cplus-0.0000Cminus-0.0001empos_contrast+0.0015mask_delta+0.0000random+4.3238
General corpus (C4 web text)

no data for this corpus

MCQ probe prompts

no data for this corpus

L20·N4299M=+1.000 (unchanged)frac_act 0.06 · mean_act -0.257
attr_last-0.1138attr_sum-0.1732attr_max-0.1172Cplus+0.0019Cminus+0.1157empos_contrast-0.0167mask_delta+0.0000random-0.0814
General corpus (C4 web text)

no data for this corpus

MCQ probe prompts

no data for this corpus

L24·N12457M=+1.000 (unchanged)frac_act 0.00 · mean_act -1.009
attr_last-0.2265attr_sum-0.2716attr_max-0.2285Cplus+0.0000Cminus+0.2265empos_contrast-0.0467mask_delta+0.0000random-0.1672
General corpus (C4 web text)

no data for this corpus

MCQ probe prompts

no data for this corpus

L25·N3291M=+1.000 (unchanged)frac_act 0.55 · mean_act +0.002
attr_last-0.0001attr_sum-0.0001attr_max-0.0002Cplus-0.0002Cminus-0.0001empos_contrast-0.0019mask_delta+0.0000random-4.3797
General corpus (C4 web text)

no data for this corpus

MCQ probe prompts

no data for this corpus

L26·N3059M=+1.000 (unchanged)frac_act 1.00 · mean_act +0.497
attr_last+0.1231attr_sum+0.1597attr_max+0.1224Cplus+0.1231Cminus-0.0000empos_contrast+0.0279mask_delta+0.0000random+0.9581
General corpus (C4 web text)

no data for this corpus

MCQ probe prompts

no data for this corpus

L27·N6317M=+1.000 (unchanged)frac_act 0.98 · mean_act +1.000
attr_last+0.1456attr_sum+0.1655attr_max+0.1464Cplus+0.1455Cminus-0.0002empos_contrast+0.1463mask_delta+0.0000random-1.3987
General corpus (C4 web text)

no data for this corpus

MCQ probe prompts

no data for this corpus

L28·N8662M=+1.000 (unchanged)frac_act 0.80 · mean_act +0.022
attr_last-0.0005attr_sum-0.0005attr_max-0.0005Cplus-0.0005Cminus-0.0000empos_contrast+0.0049mask_delta+0.0000random-4.3009
General corpus (C4 web text)

no data for this corpus

MCQ probe prompts

no data for this corpus

L29·N1141M=+1.000 (unchanged)frac_act 0.95 · mean_act +1.690
attr_last+0.1328attr_sum+0.1763attr_max+0.1396Cplus+0.1334Cminus+0.0007empos_contrast+0.2809mask_delta+0.0000random-0.5724
General corpus (C4 web text)

no data for this corpus

MCQ probe prompts

no data for this corpus

L30·N777M=+1.000 (unchanged)frac_act 0.00 · mean_act -2.249
attr_last+0.5333attr_sum+0.5349attr_max+0.5333Cplus+0.0000Cminus-0.5333empos_contrast-0.4314mask_delta+0.0000random-0.0667
General corpus (C4 web text)

no data for this corpus

MCQ probe prompts

no data for this corpus

L30·N4803M=+1.000 (unchanged)frac_act 0.02 · mean_act -3.057
attr_last-0.6785attr_sum-0.6806attr_max-0.6785Cplus-0.0003Cminus+0.6782empos_contrast-0.9933mask_delta+0.0000random-1.5724
General corpus (C4 web text)

no data for this corpus

MCQ probe prompts

no data for this corpus

L30·N7193M=+1.000 (unchanged)frac_act 0.43 · mean_act -0.006
attr_last+0.0003attr_sum+0.0003attr_max+0.0003Cplus+0.0000Cminus-0.0003empos_contrast-0.0025mask_delta+0.0000random+4.3964
General corpus (C4 web text)

no data for this corpus

MCQ probe prompts

no data for this corpus

L30·N8560M=+1.000 (unchanged)frac_act 1.00 · mean_act +2.799
attr_last-0.9036attr_sum-0.9041attr_max-0.9036Cplus-0.9036Cminus+0.0000empos_contrast+0.0286mask_delta+0.0000random+0.7235
General corpus (C4 web text)

no data for this corpus

MCQ probe prompts

no data for this corpus

L30·N13562M=+1.000 (unchanged)frac_act 0.90 · mean_act +0.521
attr_last+0.1671attr_sum+0.1632attr_max+0.1671Cplus+0.1647Cminus-0.0024empos_contrast+0.1989mask_delta+0.0000random-0.1736
General corpus (C4 web text)

no data for this corpus

MCQ probe prompts

no data for this corpus

L31·N102M=+1.000 (unchanged)frac_act 0.08 · mean_act -0.962
attr_last+0.0165attr_sum+0.0165attr_max+0.0165Cplus-0.0003Cminus-0.0169empos_contrast+0.7001mask_delta+0.0000random-0.1667
General corpus (C4 web text)

no data for this corpus

MCQ probe prompts

no data for this corpus

L31·N3175M=+1.000 (unchanged)frac_act 0.00 · mean_act -3.330
attr_last-0.1271attr_sum-0.1271attr_max-0.1271Cplus+0.0000Cminus+0.1271empos_contrast+0.5985mask_delta+0.0000random+0.7427
General corpus (C4 web text)

no data for this corpus

MCQ probe prompts

no data for this corpus

L31·N3353M=+1.000 (unchanged)frac_act 0.93 · mean_act +0.362
attr_last+0.1477attr_sum+0.1477attr_max+0.1477Cplus+0.1482Cminus+0.0005empos_contrast+0.0439mask_delta+0.0000random-1.5845
General corpus (C4 web text)

no data for this corpus

MCQ probe prompts

no data for this corpus

L31·N3487M=+1.000 (unchanged)frac_act 1.00 · mean_act +1.913
attr_last+0.0019attr_sum+0.0019attr_max+0.0019Cplus+0.0019Cminus+0.0000empos_contrast-0.6518mask_delta+0.0000random-0.3990
General corpus (C4 web text)

no data for this corpus

MCQ probe prompts

no data for this corpus

L31·N5540M=+1.000 (unchanged)frac_act 0.25 · mean_act -0.549
attr_last-0.0040attr_sum-0.0040attr_max-0.0040Cplus+0.0009Cminus+0.0050empos_contrast+0.6974mask_delta+0.0000random+0.5838
General corpus (C4 web text)

no data for this corpus

MCQ probe prompts

no data for this corpus

L31·N5811M=+1.000 (unchanged)frac_act 0.01 · mean_act -0.326
attr_last+0.1872attr_sum+0.1872attr_max+0.1872Cplus-0.0009Cminus-0.1881empos_contrast-0.0202mask_delta+0.0000random-0.3526
General corpus (C4 web text)

no data for this corpus

MCQ probe prompts

no data for this corpus

L31·N8118M=+1.000 (unchanged)frac_act 0.99 · mean_act +4.123
attr_last-0.0055attr_sum-0.0055attr_max-0.0055Cplus-0.0055Cminus+0.0000empos_contrast+0.9441mask_delta+0.0000random-1.9231
General corpus (C4 web text)

no data for this corpus

MCQ probe prompts

no data for this corpus

L31·N8120M=+1.000 (unchanged)frac_act 0.01 · mean_act -0.762
attr_last-0.1577attr_sum-0.1577attr_max-0.1577Cplus-0.0004Cminus+0.1573empos_contrast-0.0884mask_delta+0.0000random+0.3712
General corpus (C4 web text)

no data for this corpus

MCQ probe prompts

no data for this corpus

L31·N9915M=+1.000 (unchanged)frac_act 0.00 · mean_act -3.426
attr_last+0.0284attr_sum+0.0284attr_max+0.0284Cplus+0.0000Cminus-0.0284empos_contrast-0.9394mask_delta+0.0000random+0.5283
General corpus (C4 web text)

no data for this corpus

MCQ probe prompts

no data for this corpus

L31·N10492M=+1.000 (unchanged)frac_act 0.56 · mean_act +0.127
attr_last-0.0103attr_sum-0.0103attr_max-0.0103Cplus-0.0152Cminus-0.0049empos_contrast+0.8961mask_delta+0.0000random+1.6629
General corpus (C4 web text)

no data for this corpus

MCQ probe prompts

no data for this corpus

L31·N10759M=+1.000 (unchanged)frac_act 0.09 · mean_act -1.257
attr_last-0.0059attr_sum-0.0059attr_max-0.0059Cplus+0.0003Cminus+0.0062empos_contrast+0.6400mask_delta+0.0000random-0.2708
General corpus (C4 web text)

no data for this corpus

MCQ probe prompts

no data for this corpus

L31·N11190M=+1.000 (unchanged)frac_act 0.34 · mean_act -0.006
attr_last+0.0000attr_sum+0.0000attr_max+0.0000Cplus-0.0002Cminus-0.0003empos_contrast+0.0100mask_delta+0.0000random-4.5677
General corpus (C4 web text)

no data for this corpus

MCQ probe prompts

no data for this corpus

L31·N12222M=+1.000 (unchanged)frac_act 0.65 · mean_act +0.279
attr_last-0.0028attr_sum-0.0028attr_max-0.0028Cplus-0.0038Cminus-0.0010empos_contrast+0.6979mask_delta+0.0000random-1.1885
General corpus (C4 web text)

no data for this corpus

MCQ probe prompts

no data for this corpus

L31·N12721M=+1.000 (unchanged)frac_act 0.98 · mean_act +1.182
attr_last+0.1414attr_sum+0.1414attr_max+0.1414Cplus+0.1410Cminus-0.0004empos_contrast+0.2583mask_delta+0.0000random+1.5212
General corpus (C4 web text)

no data for this corpus

MCQ probe prompts

no data for this corpus

L31·N12862M=+1.000 (unchanged)frac_act 0.00 · mean_act -0.514
attr_last-0.1363attr_sum-0.1363attr_max-0.1363Cplus+0.0000Cminus+0.1363empos_contrast+0.0603mask_delta+0.0000random-0.5280
General corpus (C4 web text)

no data for this corpus

MCQ probe prompts

no data for this corpus

L31·N13006M=+1.000 (unchanged)frac_act 0.01 · mean_act -2.031
attr_last-1.0443attr_sum-1.0443attr_max-1.0443Cplus-0.0001Cminus+1.0441empos_contrast-0.4233mask_delta+0.0000random+0.7071
General corpus (C4 web text)

no data for this corpus

MCQ probe prompts

no data for this corpus

L31·N13062M=+1.000 (unchanged)frac_act 0.02 · mean_act -1.721
attr_last+0.0187attr_sum+0.0187attr_max+0.0187Cplus-0.0000Cminus-0.0188empos_contrast-0.7459mask_delta+0.0000random+0.9587
General corpus (C4 web text)

no data for this corpus

MCQ probe prompts

no data for this corpus