This story is one of 1,000 stories generated for the emotion upset. During extraction, it was fed through Gemma4-31B and its hidden state activations were captured at 11 layers.
The mean activation across all 1,000 upset stories, after denoising with neutral dialogue baselines, produces the upset emotion vector -- a direction in the model's 5,376-dimensional representation space.
Tokens promoted/suppressed when the upset vector is projected through the unembedding matrix.
S | 0.395 |
š | 0.303 |
ę ę³ | 0.246 |
ness | 0.238 |
unable | 0.238 |
de | -0.533 |
la | -0.457 |
l | -0.223 |
š | -0.215 |
š¤© | -0.210 |