Sarah slumped onto the threadbare couch, the cushions offering little comfort. Another rejection email. Another day of applying for jobs that seemed just out of reach. The dim living room, usually buzzing with the energy of her roommates, felt cavernous tonight. Her shoulders ached, a familiar tightness that settled in the space between her shoulder blades after particularly long days staring at a computer screen. She reached for the remote, the click echoing in the sudden silence as the television flickered to life. A commercial for a university popped up, the polished faces and bright smiles grating on her nerves.

The next morning, bleary-eyed and fueled by lukewarm coffee, she aimlessly scrolled through her phone. A news article caught her eye – "Best College Application Essays of the Decade." Curiosity, or maybe just a desperate need for a distraction, pulled her in. As she scanned the list, her heart leaped into her throat. There it was: her essay, her words, staring back at her. The essay she'd poured her heart into, the essay that had secured her spot at the university she’d dreamed of attending. Now, years later, it felt like a stranger's story.

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