The pizza box sat unopened on the kitchen counter, congealing in the humid afternoon heat. Sarah had ordered it hours ago, a desperate attempt to fill the gaping hole in her stomach, and maybe, just maybe, the one in her chest. Now, the thought of lifting the lid felt like scaling a mountain. The silence in the apartment was deafening, punctuated only by the rhythmic tick of the grandfather clock her grandmother had given her. It sounded slower than usual. Slower, like it was mocking her.

Her phone vibrated again, another text from Emily: "Packing is HELL! Coffee later?" Sarah stared at the screen, the cheerful emojis mocking the despair that clung to her like a wet blanket. A wave of nausea rolled over her. Coffee? Outside? Everything felt like an insurmountable effort.

She sank back onto the couch, the cushions offering little comfort. The apartment, usually a haven of cozy familiarity, now felt like a prison. Sunlight streamed through the blinds, illuminating dust motes dancing in the air, a silent, swirling testament to the chaos of her current state.

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