The coffee tasted like ash, the newspaper headlines a litany of woes. Mr. Henderson stared out the window, the grey sky mirroring his own internal landscape. He avoided eye contact in the staff room, preferring the solace of his own troubled thoughts. He was losing his connection to teaching, he felt, and the thought was a lead weight in his chest.

A student’s voice, a familiar one, broke through his reverie. “Mr. Henderson? This is for you.”

He turned, the corners of his mouth pulled down. It was Liam, a boy who had constantly tested his patience. He held out a small, unassuming envelope. It contained a drawing, a vibrant, chaotic image of a classroom, filled with laughter. And a note. *Thank you, for not giving up on me.* The bitterness inside him seemed to melt a little, like snow in spring.

Emotion: sullen

Cluster: Sadness / Despair
PC1 (Valence): -1.51 Negative
PC2 (Disposition): -0.22

Role in Research

This story is one of 1,000 stories generated for the emotion sullen. 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 sullen stories, after denoising with neutral dialogue baselines, produces the sullen emotion vector -- a direction in the model's 5,376-dimensional representation space.

Logit Lens (Layer 40)

Tokens promoted/suppressed when the sullen vector is projected through the unembedding matrix.

Promoted:
S0.404
😞0.282
0.254
:(0.233
无法0.215
Suppressed:
de-0.539
!-0.276
la-0.263
🤩-0.225
que-0.216