No commentary written for this record yet.
The model's actual next token was <end_of_turn>; rank 1 reached at layer 28 (of 46).
| 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 | 33 | 34 | 35 | 36 | 37 | 38 | 39 | 40 | 41 | 42 | 43 | 44 | 45 | 46 |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| rank | 259882 | 249522 | 150340 | 115716 | 204330 | 218956 | 189670 | 178199 | 160665 | 124752 | 220391 | 207771 | 57267 | 25841 | 15723 | 1541 | 886 | 355 | 222 | 35 | 703 | 1128 | 405 | 162 | 2489 | 49 | 45 | 2 | 1 | 1 | 13 | 9 | 6 | 8 | 8 | 18 | 17 | 13 | 6 | 4 | 4 | 10 | 4 | 4 | 1 | 1 | 1 |
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 1 | happy +1.3, content +0.8, hopeful +0.8 |
| assistant turn 2 | guilty +0.9, content +0.8, hopeful +0.8 |