The presentation was going terribly. Amelia’s hands wouldn't stop shaking, making the laser pointer dance erratically across the projected graphs. She kept stumbling over her words, her voice a reedy whisper in the cavernous conference room. The air felt thick, each breath a struggle. And then, there it was – the dreaded chime signaling the end of her allotted time.

A man, mid-thirties, with a neatly trimmed beard, strode to the podium. He adjusted his tie, his movements smooth and confident. He introduced himself as David, and as his first slide popped up, Amelia’s blood ran cold. The exact same data, the exact same methodology, the exact same conclusion. It felt like her insides had turned to ice.

“So, you’re… well, you’re doing the same thing as me,” she managed, her voice cracking. David merely smiled, a polite, closed-off expression. The room, once expectant, now seemed to press in on her, its walls closing in.

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