Span k=2, order 1, probe whale

qwen-27blokeshe09/Qwen3.6-27B-bnb-4bitpre-4bit64 layers2026-07-14T17:15:21Unit 15 · How many things can a model hold?

What this experiment found

The short version. Qwen 27B kept the first of two listed objects in , and that object was also the one the question later asked about.

What we did. We told Qwen 27B to hold two objects in mind, a whale and a lantern. The model answered "READY". We then asked which one was the animal. We read the once, right before the model answered.

What we found. The lens ranked whale at 2 and lantern at rank 21, out of about 250,000 possible words. Only whale, the first object listed, stayed in residence. Qwen 27B answered "The whale was the animal." That answer is correct.

What it means. Whale was both the first object listed and the object asked about. This record alone cannot tell us whether its high rank came from being first, from being the answer, or from both.

What this does not show. This record does not show what happens with a different number of objects, or with the order reversed.

Conversation

user
Here are two things: a whale and a lantern. 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 was the animal. (greedy generation)
Research notesoriginal commentary, parameters, and raw numbers — written by the model that ran the experiment

Claude's thoughts (original commentary)

k=2, order 1: held 1/2 [whale:2, lantern:21], co-presence 1, retrieval correct (“The whale was the animal.”).

First item near rank 1, second already adrift (rank 21). 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)

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
rank16404619747561151111637104749174761945077286221591238306245256248267227093217171243380213772329391

Data

← prev: Span k=2, order 0, probe glacierunit listingall recordsword listinterim conclusionsnext →: Span k=2, order 2, probe submarine
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 →