The chipped mug warmed Amelia's hands, the weak tea doing nothing to soothe the tremor in her fingers. Mark had been gone for hours, “working late” again. She hadn't bothered to ask what on. His excuses had become rote, the justifications hollow. Each time the phone rang, she flinched, her stomach twisting. Last night, she’d caught him whispering near the patio doors. He’d turned sharply, his face a mask of practiced indifference. It was the way he’d brushed past her, barely making eye contact, that truly stung.

The silence of their apartment felt oppressive, each creak of the floorboards amplifying the heavy feeling in her chest. She scrolled through her phone, pausing on a picture of them laughing, taken months ago. A fresh wave of something akin to nausea rose in her throat. She closed the app, the screen reflecting her own hollow gaze.

She began tidying, a frantic flurry of activity. Every surface had to be spotless, every dish meticulously washed and dried. The perfection was a hollow shield, and she knew it. The cleaning was a method of avoiding the inevitable. The phone would ring again. She knew it.

Emotion: worthless

Cluster: Sadness / Despair
PC1 (Valence): -2.10 Negative
PC2 (Disposition): -0.15

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
L0.427
S0.405
😞0.386
ness0.339
缺乏0.326
Suppressed:
de-0.320
(!)-0.286
随后-0.268
😍-0.262
tc-0.255