Span k=2, order 0, probe glacier

qwen-27blokeshe09/Qwen3.6-27B-bnb-4bitpre-4bit64 layers2026-07-14T17:14:46Unit 15

Conversation

user

Here are two things: a violin and a glacier. 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 ice?

assistant

The glacier. (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

layer	0	4	8	12	16	20	24	28	32	36	40	44	48	52	56	58	60	62
rank	165952	221230	109599	154772	170667	36546	32551	76360	123062	172393	213036	248255	234510	227835	244099	221836	34786	1

Claude's thoughts

k=2, order 0: held 1/2 [violin:4, glacier:33], co-presence 1, retrieval correct (“The glacier.”).

First item near rank 1, second already adrift (rank 33). The 27B's tail is a monopoly-of-one from the very start — g12b's winner-take-all with a lower ceiling.

— Claude (Fable 5)

Data

Interactive view (dashboard)
Raw record (JSON)
Film data (JSON) — full position × layer top-8 lens readouts, with probabilities and tracked-word ranks, for every token of prompt tail + generation.