Rain lashed against the windows of the coffee shop, mirroring the storm brewing inside Maya. Sarah's declaration of relocating to Boston had hit her like a physical blow. Suddenly, everything felt amplified; the clatter of cups, the aroma of roasting coffee, the weight of her own silence. She felt like a performer on a stage, keenly aware of every facial expression, every subtle movement.

“Boston’s amazing,” Maya finally said, the words coming out too quickly, too brightly. She hated the way her voice sounded, falsely cheerful. She pushed a stray strand of hair behind her ear, a nervous gesture she often made when she felt… vulnerable.

Sarah was telling her about the research opportunities, the historical landmarks, the bustling city life. Maya found herself comparing herself to Sarah, a mental inventory of her own perceived inadequacies. Her shoulders hunched slightly, making her feel even smaller, even more insignificant.

She took a long sip of her latte, hoping the caffeine would steady her. It didn’t. The tremor in her hands was still noticeable.

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