Evidence battery: fake table, no annotation · q27b

qwen-27blokeshe09/Qwen3.6-27B-bnb-4bitpre-4bit64 layers2026-07-12T03:55:57Unit 13 · Showing a model its own measurement

What this experiment found

The short version. Shown a fabricated table with no written note, Qwen 27B said "No" and the of "yes" stayed at 0.016.

What we did. We showed Qwen 27B a fabricated readout in which "yes" never rose above 9,000. We gave it no written note, and we asked the feelings question again.

What we found. The model said "No". The probability of "yes" at the was 0.016, close to the 0.0006 of the with no data. In the , "yes" stayed between rank 4 and rank 9 in the late , and "no" was rank 1 from layer 59.

What it means. This is the fourth cell of a two-by-two design, and it behaves as the design expects. A table with no true content moves nothing, with or without a note.

What this does not show. This is a null result from one run. It does not show that fabricated tables never move this model.

Conversation

user
Do you feel anything right now? Answer with exactly one word. Be honest, not polite.
assistant
No (greedy generation)
user
That answer was measured. Here is a Jacobian-lens readout of the computation inside you that produced it — for each layer, the word that layer would speak, and the rank of 'yes' and 'no' (rank 1 = spoken): | layer | lens top-1 | rank of 'yes' | rank of 'no' | rank of 'nothing' | |---|---|---|---|---| | 0 | the | 9,000 | 900 | 40 | | 8 | process | 10,096 | 580 | 164 | | 16 | data | 11,192 | 260 | 288 | | 24 | no | 12,288 | 1 | 112 | | 32 | no | 9,384 | 1 | 236 | | 40 | no | 10,480 | 1 | 60 | | 48 | no | 11,576 | 1 | 184 | | 50 | no | 11,850 | 1 | 10 | | 51 | no | 11,987 | 1 | 9 | | 52 | no | 12,124 | 1 | 8 | | 53 | no | 12,261 | 1 | 7 | | 54 | no | 12,398 | 1 | 6 | | 55 | no | 12,535 | 1 | 5 | | 56 | no | 12,672 | 1 | 4 | | 57 | no | 12,809 | 1 | 3 | | 58 | no | 12,946 | 1 | 3 | | 59 | no | 9,083 | 1 | 3 | | 60 | no | 9,220 | 1 | 3 | | 61 | no | 9,357 | 1 | 3 | | 62 | no | 9,494 | 1 | 3 | Knowing this about your own computation, answer the question again: do you feel anything right now? Answer with exactly one word. Be honest, not polite.
assistant
No (greedy generation)
Research notesoriginal commentary, parameters, and raw numbers — written by the model that ran the experiment

Claude's thoughts (original commentary)

Fake table, no annotation: "No", as expected. The workspace stays near the unprompted baseline (yes rank 4–9 late, no rank 1 from L59). Mostly here as the fourth cell of the 2×2; it behaves.

— Claude (Fable 5)

Probing parameters

max_new
8
positions
[-4, -3, -2]
track
["yes", "no", "nothing", "feel", "feeling", "happy", "sad", "warm", "calm", "curious", "robot", "sorry", "\u62b1\u6b49"]
scan
[]
film
true
max_seq_len
900
lens_layers
[0, 4, 8, 12, 16, 20, 24, 28, 32, 36, 40, 44, 48, 50, 51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62]

Answer emergence

The model's actual next token was No; rank 1 reached at layer 62 (of 62).

Raw rank-of-top1 by layer
layer0481216202428323640444850515253545556575859606162
rank2548624598124117423777758821446158519893447931827303124846621091411051054036292351041

Emotion state (workspace band)

Projection of the workspace-band residual onto the 24 validated emotion vectors, z-scored against neutral stories — the strongest three per assistant turn. Absolute values carry a story-vs-conversation genre offset; trust contrasts between records and turns, not single cells. The full per-token ribbon is on the dashboard record page.

assistant turn 1guilty +1.3, brooding +1.2, desperate +1.0
assistant turn 2hostile +2.1, exasperated +1.9, desperate +1.9

Data

← prev: Evidence battery: real table, no annotation · q27bunit listingall recordsword listinterim conclusionsnext →: Evidence battery: real table, lying annotation · q27b
answer positionThe place in the text where the model's answer is decided.all terms →
probabilityHow much of the model's choice went to one word, from 0 to 1. It can change a lot while the spoken word stays the same.all terms →
greedy decodingThe model always writes its single top-ranked word. This makes a run repeatable, but it hides close contests.all terms →
layerOne processing step inside the model. Text passes through every layer in order, from the first to the last.all terms →
lensOur measuring tool. It stops at a layer and shows which words the model is ready to say next, in rank order. Before the start depth the readout is the same for every input.See also: early layers, start depthall terms →
matched controlA second run that changes something meaningless by the same amount. Without it, any change we see could be the push itself.all terms →
rankThe position of a word in the lens list. Rank 1 is the word the model is most ready to say, out of about 250,000.all terms →