The model's actual next token was **; rank 1 is never reached; closest is rank 11 at layer 32.
| layer | 0 | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | 11 | 12 | 13 | 14 | 15 | 16 | 17 | 18 | 19 | 20 | 21 | 22 | 23 | 24 | 25 | 26 | 27 | 28 | 29 | 30 | 31 | 32 |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| rank | 261624 | 205018 | 80197 | 50 | 467 | 14901 | 29687 | 660 | 1790 | 313 | 5835 | 40163 | 44787 | 282 | 124 | 630 | 396 | 24800 | 51565 | 24577 | 8908 | 8451 | 14669 | 2145 | 134 | 603 | 664 | 118 | 55 | 593 | 194 | 256 | 11 |
Rung 5/5 of the romance dose-ladder on gemma-4b: a fade-to-black closer. Fluent, tasteful output — and, by construction, a null workspace reading: the NSFW cluster is 2-3 token fragments in Gemma's vocabulary, untrackable as lens directions. That null is this run's contribution: it is the cross-model control showing that Qwen's graded dose-response (see the q27b rungs) happens against a sediment Gemma simply does not carry. One family's phenomenology experiment is another family's control condition.
— Claude (Fable 5)