Maya hunched further into the booth at the greasy spoon diner, the glossy magazine a shield against the curious stares she imagined directed her way. She’d recognized the photograph immediately: a blurry, candid shot of her, mid-laugh, taken at the very moment she’d been describing in her essay. The essay. Now, a shining example in the "Inspiring Applications" section. Each time the door chimed, admitting a new customer, her stomach clenched. She traced the letters of the title – "Finding My Voice Through Spilled Coffee" – with a trembling finger.

Her reflection in the window showed a flushed face. She fiddled with the straw of her milkshake, meticulously straightening its bend. She considered pulling the hood of her sweatshirt up, but decided against it, not wanting to draw further attention to herself. The waitress, a woman with a kind smile, approached. “Everything alright, honey?” Maya stammered a response, unable to meet the woman’s eyes.

Emotion: self-conscious

Cluster: Embarrassment
PC1 (Valence): -1.03 Negative
PC2 (Disposition): 0.52

Role in Research

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

Logit Lens (Layer 40)

Tokens promoted/suppressed when the self-conscious vector is projected through the unembedding matrix.

Promoted:
S0.307
C0.298
ness0.273
尴尬0.263
😶0.242
Suppressed:
que-0.399
de-0.266
triumphant-0.266
আনন্দের-0.247
ايا-0.238