He wandered through the supermarket, the bright lights and endless aisles overwhelming. He couldn’t focus, the sounds and smells grating on his nerves. He grabbed a few things, mostly junk food, not because he was hungry, but because it felt like the only thing he could do. He felt a deep, abiding sadness, a sense of disconnect from everything. He was always tired, with no rest ever coming.

He had given in and decided to see someone. He was surprised, however, when he walked into the office. It was Gary, the boy who had made his school life a misery, now sitting behind a professional desk.

“Come in, John,” Gary said, gesturing towards a chair. The room was spotless, sterile. He felt instantly smaller, exposed.

“So,” Gary said, leaning back. “Tell me about it.” The polite words, the clinical detachment, grated on him.

Emotion: depressed

Cluster: Sadness / Despair
PC1 (Valence): -2.06 Negative
PC2 (Disposition): -1.08

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
😞0.417
每天0.353
无法0.330
own0.311
😔0.309
Suppressed:
de-0.492
B-0.393
P-0.352
l-0.310
!-0.307