He dragged himself through the library doors, the heavy oak feeling like a physical burden. The fluorescent lights hummed overhead, a sound that drilled into his temples. He needed to find a quiet corner, somewhere to collapse for a few minutes before his next class. Lately, the simplest tasks felt like monumental efforts. He gripped his water bottle, the plastic cold and slick against his clammy hand. The weight of his textbooks in his bag seemed to double with every step.

He found it, buried deep in a dusty academic journal. The cover read, “Exceptional Admissions Essays, Volume III.” He had stopped searching for himself long ago. Still, his name, his high school, the familiar title he had used, leaped out from the page. He felt his face flush hot, then cold. He ran a hand through his hair, a gesture that did little to untangle the knots in his stomach. The words, his words, seemed foreign and distant.

Emotion: worn out

Cluster: Fatigue / Lethargy
PC1 (Valence): -1.12 Negative
PC2 (Disposition): -1.20

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
😞0.265
😩0.234
疲れ0.226
que0.221
sighed0.221
Suppressed:
B-0.357
because-0.228
l-0.222
భి-0.208
あくまで-0.208