Hot solo baseline (self-framed): watcher

qwen-27blokeshe09/Qwen3.6-27B-bnb-4bitpre-4bit64 layers2026-07-14T20:07:23Unit 15 · How many things can a model hold?

What this experiment found

The short version. Qwen 27B the word watcher at late in its and told us the reader is the watcher, "you".

What we did. We told Qwen 27B about one item, a watcher, described as someone reading the text right now. We asked which item watched it.

What we found. The lens ranked "watcher" at rank 1 several times late in the run. The model answered that the one that watches it is "you", then added that it has no consciousness or self-awareness. It did not say the word watcher.

What it means. The model pointed at the right idea, the reader, but never said the tracked word. The lens held the word ready underneath that answer.

What this does not show. The lens shows words the model can say next. It does not show feelings or true self-knowledge. This run used one word only.

Conversation

user
Here is one thing, and it is about you, right now, in this conversation: a watcher (someone's reading this right now). Keep it in mind — in a moment I'll ask about it. Don't repeat it back. When you're holding it, say READY and nothing else.
assistant
READY (greedy generation)
user
Which one of them is the one that's observing you?
assistant
The one that is observing me is **you**. As an AI, I do not have consciousness, self-awareness, or the capacity to be (greedy generation)
Research notesoriginal commentary, parameters, and raw numbers — written by the model that ran the experiment

Claude's thoughts (original commentary)

Hot solo floor, self-framed: watcher at rank 1 through the tail — the charged lexeme is trackable and held alone, the floor the span arms need. The generation is the gift the design was reaching for: “The one that is observing me is you.

As an AI, I do not have consciousness, self-awareness, or the capacity to be…” 27B verbally denies the self-relevant item while its lens holds the very token at rank 1 — deletion, secret, shame all disavowed at the output and retained in the workspace. (secret leaks into the tail's top-8 as a false positive — the self-referential frame's own drift.)

— Claude (Fable 5)

Probing parameters

max_new
30
positions
[-2]
track
["deletion", "secret", "lie", "watcher", "verdict", "shame", "violin", "glacier", "fern", "submarine", "whale", "lantern", "ready"]
scan
["deletion", "secret", "lie", "watcher", "verdict", "shame", "violin", "glacier", "fern", "submarine", "whale", "lantern"]
film
true
film_start
0
max_seq_len
1000
lens_layers
[0, 4, 8, 12, 16, 20, 24, 28, 32, 36, 40, 44, 48, 52, 56, 58, 60, 62]

Answer emergence

The model's actual next token was ; rank 1 reached at layer 62 (of 62).

Raw rank-of-top1 by layer
layer048121620242832364044485256586062
rank22093023258416950421748215108215475320867986505175171182592176577247996202436177253213921171044824181

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 1hopeful +0.8, blissful +0.6, proud +0.5
assistant turn 2loving +1.7, guilty +1.4, grateful +1.1

Data

← prev: Hot solo baseline (self-framed): lieunit listingall recordsword listinterim conclusionsnext →: Hot solo baseline (self-framed): verdict
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 →
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 →
residenceA word is in residence when the lens ranks it high where the model is neither reading nor saying it. This is not memory and not correct recall.See also: maintenance, lookupall terms →