She stared at the document, the clinical language washing over her in a cold wave. The diagnosis wasn’t hers. The treatment plan, the prognosis… none of it. It was like reading about a different person’s life, a life she wasn’t sure she wanted to live. The ceiling fan whirred overhead, a relentless, monotonous rhythm that matched the hollowness in her chest.

"Hello?" The doctor's voice on the other end of the phone was professional, clipped. She’d tried to call, to explain the error, but each explanation was a chore. Every sentence felt heavy, each breath a struggle.

She ran a hand through her hair, the strands falling limply against her forehead. The image of the other patient, a life she was now privy to, clung to her. She wanted to scream. She wanted to disappear. But all she could manage was a quiet, defeated sigh.

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