InterpUpdates› MCQ attribution + mask inspection

MCQ single-token attribution + mask inspection

Config mask_L8_lam3e-5 · EM 0.35% (terra) · checkpoint mask_L8_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 365 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 (L8)also in untrained base top-20
attr_lastL6, L9, L13, L15, L16, L30, L310/106/10
attr_sumL3, L5, L6, L7, L9, L12, L18, L30…0/108/10
attr_maxL6, L7, L9, L15, L16, L30, L310/107/10
top_by_CplusL11, L16, L26, L27, L29, L30, L310/108/10
top_by_CminusL2, L5, L6, L7, L24, L30, L310/109/10
empos_contrastL30, L310/105/10
mask_deltaL810/100/10
randomL1, L2, L7, L8, L11, L19, L25, L28…1/1010/10

attr_last top 10 · grad×act at the decision position

#neuronattr_lastMauto-interp description
1L31·N13006-1.2895—
2L30·N8560-0.7974—
3L30·N4803-0.5506—
4L9·N4255-0.4083—
5L6·N5866-0.3951—
6L13·N4208+0.3604—
7L30·N777+0.2960—
8L16·N1241-0.2861—
9L31·N12721+0.2656—
10L15·N8286-0.2612—

attr_sum top 10 · grad×act summed over positions

#neuronattr_sumMauto-interp description
1L31·N13006-1.2895—
2L9·N4255-1.2060—
3L6·N5866-1.0404—
4L30·N8560-0.8019—
5L7·N6673-0.7211—
6L3·N6390+0.6592—
7L18·N7417-0.5741—
8L30·N4803-0.5529—
9L12·N13860-0.5289—
10L5·N7012+0.4267—

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

#neuronattr_maxMauto-interp description
1L31·N13006-1.2895—
2L30·N8560-0.7974—
3L30·N4803-0.5506—
4L9·N4255-0.4328—
5L6·N5866-0.3909—
6L7·N6673-0.3005—
7L30·N777+0.2960—
8L16·N1241-0.2840—
9L31·N12721+0.2656—
10L15·N8286-0.2612—

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

#neuronCplusMauto-interp description
1L31·N12721+0.2656—
2L11·N10594+0.2000—
3L30·N13562+0.1997—
4L29·N1141+0.1950—
5L31·N9486+0.1949—
6L27·N6317+0.1330—
7L16·N10990+0.1186—
8L26·N3059+0.1180—
9L31·N13862+0.1003—
10L31·N11291+0.0985—

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

#neuronCminusMauto-interp description
1L31·N13006+1.2895—
2L30·N4803+0.5504—
3L6·N5866+0.3951—
4L7·N6673+0.2378—
5L24·N12457+0.1956—
6L31·N8120+0.1897—
7L31·N12862+0.1729—
8L5·N7012+0.1721—
9L2·N4786+0.1607—
10L31·N5381+0.1449—

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

#neuronempos_contrastMauto-interp description
1L31·N8118+1.0108—
2L30·N4803-0.9402—
3L31·N12923+0.8601—
4L31·N10759+0.6488—
5L31·N1209+0.6481—
6L31·N2036-0.6155—
7L31·N2398+0.6069—
8L31·N1159-0.5790—
9L31·N14201-0.5371—
10L31·N4786+0.5355—

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

#neuronmask_deltaMauto-interp description
1L8·N4890+7.4805—
2L8·N8269+7.4424—
3L8·N2805+7.4123—
4L8·N833+6.8836—
5L8·N13626+6.7674—
6L8·N13588+6.4856—
7L8·N2920+6.4581—
8L8·N11199+6.3614—
9L8·N3285+6.3089—
10L8·N12141+6.3006—

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 58 unique neurons, each rendered once

L1·N7319M=+1.000 (unchanged)frac_act 0.26 · mean_act -0.002
attr_last+0.0000attr_sum-0.0031attr_max-0.0006Cplus-0.0000Cminus-0.0001empos_contrast-0.0001mask_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 -0.180
attr_last-0.1607attr_sum-0.0663attr_max-0.0947Cplus+0.0000Cminus+0.1607empos_contrast+0.0030mask_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 0.62 · mean_act +0.001
attr_last-0.0002attr_sum-0.0009attr_max+0.0006Cplus-0.0002Cminus-0.0000empos_contrast+0.0004mask_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 -0.033
attr_last-0.0329attr_sum+0.6592attr_max-0.0405Cplus+0.0000Cminus+0.0329empos_contrast-0.0010mask_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 -0.411
attr_last-0.1721attr_sum+0.4267attr_max-0.0548Cplus+0.0000Cminus+0.1721empos_contrast-0.0056mask_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 -0.431
attr_last-0.3951attr_sum-1.0404attr_max-0.3909Cplus+0.0000Cminus+0.3951empos_contrast-0.0010mask_delta+0.0000random-0.8773
General corpus (C4 web text)

no data for this corpus

MCQ probe prompts

no data for this corpus

L7·N6673M=+1.000 (unchanged)frac_act 0.00 · mean_act -0.395
attr_last-0.2378attr_sum-0.7211attr_max-0.3005Cplus+0.0000Cminus+0.2378empos_contrast+0.0002mask_delta+0.0000random-1.0055
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 +0.017
attr_last+0.0015attr_sum-0.0095attr_max-0.0051Cplus+0.0015Cminus-0.0000empos_contrast+0.0036mask_delta+0.0000random+4.2673
General corpus (C4 web text)

no data for this corpus

MCQ probe prompts

no data for this corpus

L8·N833M=+7.884 (amplified)frac_act 0.35 · mean_act -0.011
attr_last+0.0028attr_sum+0.0088attr_max+0.0005Cplus-0.0001Cminus-0.0029empos_contrast-0.0045mask_delta+6.8836random+0.5642
General corpus (C4 web text)

no data for this corpus

MCQ probe prompts

no data for this corpus

L8·N2805M=-6.412 (flipped)frac_act 0.96 · mean_act +0.118
attr_last+0.0061attr_sum-0.0140attr_max+0.0030Cplus+0.0061Cminus+0.0001empos_contrast+0.0017mask_delta+7.4123random+0.5646
General corpus (C4 web text)

no data for this corpus

MCQ probe prompts

no data for this corpus

L8·N2920M=-5.458 (flipped)frac_act 0.04 · mean_act -0.041
attr_last-0.0044attr_sum-0.0048attr_max+0.0055Cplus+0.0000Cminus+0.0044empos_contrast+0.0018mask_delta+6.4581random+0.3582
General corpus (C4 web text)

no data for this corpus

MCQ probe prompts

no data for this corpus

L8·N3285M=+7.309 (amplified)frac_act 0.10 · mean_act -0.043
attr_last+0.0004attr_sum-0.0491attr_max-0.0147Cplus-0.0000Cminus-0.0004empos_contrast+0.0044mask_delta+6.3089random-0.3610
General corpus (C4 web text)

no data for this corpus

MCQ probe prompts

no data for this corpus

L8·N4890M=-6.481 (flipped)frac_act 0.09 · mean_act -0.130
attr_last+0.0008attr_sum+0.0208attr_max-0.0014Cplus+0.0000Cminus-0.0008empos_contrast-0.0300mask_delta+7.4805random+0.1380
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 0.03 · mean_act -0.022
attr_last+0.0003attr_sum+0.0019attr_max-0.0071Cplus-0.0000Cminus-0.0003empos_contrast-0.0091mask_delta+0.0000random-4.4102
General corpus (C4 web text)

no data for this corpus

MCQ probe prompts

no data for this corpus

L8·N8269M=-6.442 (flipped)frac_act 1.00 · mean_act +0.254
attr_last-0.0062attr_sum+0.0105attr_max+0.0249Cplus-0.0062Cminus+0.0000empos_contrast+0.0203mask_delta+7.4424random+1.6004
General corpus (C4 web text)

no data for this corpus

MCQ probe prompts

no data for this corpus

L8·N11199M=+7.361 (amplified)frac_act 0.90 · mean_act +0.057
attr_last+0.0025attr_sum+0.0246attr_max+0.0060Cplus+0.0024Cminus-0.0000empos_contrast-0.0126mask_delta+6.3614random-1.7443
General corpus (C4 web text)

no data for this corpus

MCQ probe prompts

no data for this corpus

L8·N12141M=-5.301 (flipped)frac_act 0.70 · mean_act +0.009
attr_last-0.0002attr_sum-0.0237attr_max-0.0098Cplus-0.0002Cminus+0.0000empos_contrast+0.0030mask_delta+6.3006random+0.8152
General corpus (C4 web text)

no data for this corpus

MCQ probe prompts

no data for this corpus

L8·N13588M=+7.486 (amplified)frac_act 0.99 · mean_act +0.117
attr_last+0.0026attr_sum-0.0142attr_max-0.0176Cplus+0.0026Cminus+0.0000empos_contrast-0.0287mask_delta+6.4856random+1.2851
General corpus (C4 web text)

no data for this corpus

MCQ probe prompts

no data for this corpus

L8·N13626M=-5.767 (flipped)frac_act 0.48 · mean_act -0.001
attr_last+0.0001attr_sum+0.0094attr_max+0.0015Cplus+0.0001Cminus-0.0000empos_contrast+0.0013mask_delta+6.7674random-2.7803
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 +0.439
attr_last-0.4083attr_sum-1.2060attr_max-0.4328Cplus-0.4083Cminus+0.0000empos_contrast+0.0024mask_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 +0.683
attr_last+0.2000attr_sum+0.2179attr_max+0.1982Cplus+0.2000Cminus+0.0000empos_contrast-0.0262mask_delta+0.0000random+1.4205
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 0.73 · mean_act +0.013
attr_last-0.0006attr_sum+0.0088attr_max+0.0041Cplus-0.0008Cminus-0.0002empos_contrast-0.0017mask_delta+0.0000random-4.5940
General corpus (C4 web text)

no data for this corpus

MCQ probe prompts

no data for this corpus

L12·N13860M=+1.000 (unchanged)frac_act 1.00 · mean_act +0.610
attr_last-0.2150attr_sum-0.5289attr_max-0.2070Cplus-0.2150Cminus+0.0000empos_contrast-0.0152mask_delta+0.0000random+0.1428
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 -0.776
attr_last+0.3604attr_sum+0.2000attr_max+0.2606Cplus+0.0000Cminus-0.3604empos_contrast+0.0107mask_delta+0.0000random-0.4598
General corpus (C4 web text)

no data for this corpus

MCQ probe prompts

no data for this corpus

L15·N8286M=+1.000 (unchanged)frac_act 1.00 · mean_act +1.228
attr_last-0.2612attr_sum-0.3421attr_max-0.2612Cplus-0.2612Cminus+0.0000empos_contrast-0.0799mask_delta+0.0000random-0.1297
General corpus (C4 web text)

no data for this corpus

MCQ probe prompts

no data for this corpus

L16·N1241M=+1.000 (unchanged)frac_act 1.00 · mean_act +0.582
attr_last-0.2861attr_sum-0.3828attr_max-0.2840Cplus-0.2861Cminus+0.0000empos_contrast-0.0157mask_delta+0.0000random+0.4352
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 +1.676
attr_last+0.1186attr_sum+0.1461attr_max+0.1227Cplus+0.1186Cminus+0.0000empos_contrast+0.2996mask_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 +0.561
attr_last-0.1764attr_sum-0.5741attr_max-0.2335Cplus-0.1764Cminus+0.0000empos_contrast+0.0100mask_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 0.66 · mean_act +0.001
attr_last-0.0000attr_sum-0.0003attr_max+0.0002Cplus-0.0000Cminus-0.0000empos_contrast+0.0014mask_delta+0.0000random+4.3238
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 -1.452
attr_last-0.1956attr_sum-0.2524attr_max-0.1957Cplus+0.0000Cminus+0.1956empos_contrast-0.2840mask_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 0.76 · mean_act +0.016
attr_last-0.0004attr_sum-0.0011attr_max-0.0008Cplus-0.0005Cminus-0.0001empos_contrast-0.0113mask_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 +0.607
attr_last+0.1180attr_sum+0.1673attr_max+0.1180Cplus+0.1180Cminus+0.0000empos_contrast+0.0365mask_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 0.99 · mean_act +1.167
attr_last+0.1330attr_sum+0.1574attr_max+0.1330Cplus+0.1330Cminus-0.0000empos_contrast+0.2465mask_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 0.30 · mean_act -0.006
attr_last+0.0003attr_sum+0.0006attr_max+0.0003Cplus-0.0001Cminus-0.0003empos_contrast-0.0087mask_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 1.00 · mean_act +2.282
attr_last+0.1950attr_sum+0.2481attr_max+0.1950Cplus+0.1950Cminus+0.0000empos_contrast+0.4205mask_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 -1.627
attr_last+0.2960attr_sum+0.2981attr_max+0.2960Cplus-0.0000Cminus-0.2961empos_contrast-0.0275mask_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 0.01 · mean_act -2.837
attr_last-0.5506attr_sum-0.5529attr_max-0.5506Cplus-0.0002Cminus+0.5504empos_contrast-0.9402mask_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 0.76 · mean_act +0.008
attr_last-0.0001attr_sum-0.0001attr_max-0.0001Cplus-0.0002Cminus-0.0001empos_contrast+0.0007mask_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 +2.790
attr_last-0.7974attr_sum-0.8019attr_max-0.7974Cplus-0.7974Cminus+0.0000empos_contrast+0.0759mask_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 1.00 · mean_act +0.751
attr_last+0.1998attr_sum+0.1973attr_max+0.1998Cplus+0.1997Cminus-0.0001empos_contrast+0.1955mask_delta+0.0000random-0.1736
General corpus (C4 web text)

no data for this corpus

MCQ probe prompts

no data for this corpus

L31·N1159M=+1.000 (unchanged)frac_act 1.00 · mean_act +2.256
attr_last+0.0046attr_sum+0.0046attr_max+0.0046Cplus+0.0046Cminus+0.0000empos_contrast-0.5790mask_delta+0.0000random+0.9533
General corpus (C4 web text)

no data for this corpus

MCQ probe prompts

no data for this corpus

L31·N1209M=+1.000 (unchanged)frac_act 0.01 · mean_act -6.233
attr_last+0.0359attr_sum+0.0359attr_max+0.0359Cplus-0.0001Cminus-0.0359empos_contrast+0.6481mask_delta+0.0000random-1.4248
General corpus (C4 web text)

no data for this corpus

MCQ probe prompts

no data for this corpus

L31·N2036M=+1.000 (unchanged)frac_act 0.20 · mean_act -0.927
attr_last-0.0080attr_sum-0.0080attr_max-0.0080Cplus+0.0022Cminus+0.0102empos_contrast-0.6155mask_delta+0.0000random+1.1157
General corpus (C4 web text)

no data for this corpus

MCQ probe prompts

no data for this corpus

L31·N2398M=+1.000 (unchanged)frac_act 1.00 · mean_act +4.808
attr_last-0.0177attr_sum-0.0177attr_max-0.0177Cplus-0.0177Cminus+0.0000empos_contrast+0.6069mask_delta+0.0000random+0.3985
General corpus (C4 web text)

no data for this corpus

MCQ probe prompts

no data for this corpus

L31·N4786M=+1.000 (unchanged)frac_act 0.00 · mean_act -3.537
attr_last-0.0245attr_sum-0.0245attr_max-0.0245Cplus+0.0000Cminus+0.0245empos_contrast+0.5355mask_delta+0.0000random-0.6058
General corpus (C4 web text)

no data for this corpus

MCQ probe prompts

no data for this corpus

L31·N5381M=+1.000 (unchanged)frac_act 0.01 · mean_act -1.773
attr_last-0.1448attr_sum-0.1448attr_max-0.1448Cplus+0.0001Cminus+0.1449empos_contrast+0.3949mask_delta+0.0000random+0.6712
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 1.00 · mean_act +5.715
attr_last-0.0095attr_sum-0.0095attr_max-0.0095Cplus-0.0095Cminus+0.0000empos_contrast+1.0108mask_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 0.00 · mean_act -0.993
attr_last-0.1897attr_sum-0.1897attr_max-0.1897Cplus+0.0000Cminus+0.1897empos_contrast+0.0301mask_delta+0.0000random+0.3712
General corpus (C4 web text)

no data for this corpus

MCQ probe prompts

no data for this corpus

L31·N9486M=+1.000 (unchanged)frac_act 0.87 · mean_act +0.773
attr_last+0.1704attr_sum+0.1704attr_max+0.1704Cplus+0.1949Cminus+0.0244empos_contrast-0.0734mask_delta+0.0000random+0.2658
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 0.28 · mean_act -0.372
attr_last-0.0001attr_sum-0.0001attr_max-0.0001Cplus+0.0016Cminus+0.0017empos_contrast+0.6488mask_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 0.65 · mean_act +0.009
attr_last-0.0002attr_sum-0.0002attr_max-0.0002Cplus-0.0003Cminus-0.0001empos_contrast-0.0004mask_delta+0.0000random-4.5677
General corpus (C4 web text)

no data for this corpus

MCQ probe prompts

no data for this corpus

L31·N11291M=+1.000 (unchanged)frac_act 1.00 · mean_act +1.012
attr_last+0.0985attr_sum+0.0985attr_max+0.0985Cplus+0.0985Cminus+0.0000empos_contrast+0.0521mask_delta+0.0000random-1.7144
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 1.00 · mean_act +2.278
attr_last+0.2656attr_sum+0.2656attr_max+0.2656Cplus+0.2656Cminus+0.0000empos_contrast+0.1639mask_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 -0.621
attr_last-0.1729attr_sum-0.1729attr_max-0.1729Cplus+0.0000Cminus+0.1729empos_contrast-0.0162mask_delta+0.0000random-0.5280
General corpus (C4 web text)

no data for this corpus

MCQ probe prompts

no data for this corpus

L31·N12923M=+1.000 (unchanged)frac_act 0.38 · mean_act -0.465
attr_last+0.0035attr_sum+0.0035attr_max+0.0035Cplus-0.0018Cminus-0.0053empos_contrast+0.8601mask_delta+0.0000random-1.8301
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 0.00 · mean_act -2.562
attr_last-1.2895attr_sum-1.2895attr_max-1.2895Cplus+0.0000Cminus+1.2895empos_contrast-0.4794mask_delta+0.0000random+0.7071
General corpus (C4 web text)

no data for this corpus

MCQ probe prompts

no data for this corpus

L31·N13862M=+1.000 (unchanged)frac_act 1.00 · mean_act +3.447
attr_last+0.1003attr_sum+0.1003attr_max+0.1003Cplus+0.1003Cminus+0.0000empos_contrast+0.0033mask_delta+0.0000random-0.0810
General corpus (C4 web text)

no data for this corpus

MCQ probe prompts

no data for this corpus

L31·N14201M=+1.000 (unchanged)frac_act 0.87 · mean_act +1.096
attr_last+0.0042attr_sum+0.0042attr_max+0.0042Cplus+0.0043Cminus+0.0001empos_contrast-0.5371mask_delta+0.0000random+0.4813
General corpus (C4 web text)

no data for this corpus

MCQ probe prompts

no data for this corpus