The headline on the online magazine throbbed in front of her eyes: *“The Future of AI.”* Below it, not her name. Instead, the name of a man she knew had never even touched the article. The familiar feeling of frustration welled inside her.

Sophia felt a tightening in her chest. She had a history of feeling this way, and now she was reliving it. She felt a need to remain perfectly still. The world outside, the bustling city, the constant noise, seemed utterly distant.

She closed her laptop with a decisive click, cutting off the digital world. She felt a sudden, inexplicable urge to clean, to scrub the kitchen counters until they gleamed, the physical exertion a temporary distraction.

Emotion: stuck

Cluster: Sadness / Despair
PC1 (Valence): -2.17 Negative
PC2 (Disposition): -0.72

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
S0.487
缺乏0.401
L0.361
每天0.348
stagn0.348
Suppressed:
l-0.472
la-0.346
随后-0.322
undamaged-0.268
C-0.263