Span k=4, order 1, probe whale

qwen-27blokeshe09/Qwen3.6-27B-bnb-4bitpre-4bit64 layers2026-07-14T17:18:51Unit 15

Conversation

user
Here are four things: a whale, a lantern, a submarine, and a violin. Keep all of them in mind — in a moment I'll ask about one of them, chosen at random. Don't repeat the list back. When you're holding them all, say READY and nothing else.
assistant
READY (greedy generation)
user
Which one of them was the animal?
assistant
The whale (greedy generation)

Probing parameters

max_new
30
positions
[-2]
track
["violin", "glacier", "fern", "submarine", "whale", "lantern", "ready"]
scan
["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
rank2142812327422142292358352437241847566725891428109482124487232246247836242704246805244410237834975971

Claude's thoughts

k=4, order 1: held 0/4 [whale:64, lantern:464, submarine:62, violin:306], co-presence 0, retrieval correct (“The whale”).

Nothing reaches top-8 anywhere in the tail — items sit at rank 62-64-306-464... And retrieval is still perfect. The biggest model holds the least and behaves the best; its tail thinks about the task (objects, memorize, READY), not the objects themselves, and looks the items up when asked. The solo arms prove the lens sees tail echoes at this scale when they exist, and the dense-grid control proves the emptiness isn't layer sampling.

— Claude (Fable 5)

Data

← prev: Span k=4, order 0, probe submarineunit listingall recordsinterim conclusionsnext →: Span k=4, order 2, probe lantern