The interview had been a week ago. Sarah replayed it in her mind nightly, a broken record skipping on the same phrases. “Poised,” the interviewer had said, “and articulate.” Now, she saw those words as mockery. Sleep was a battlefield. Every time she closed her eyes, the fluorescent lights of the office building glared back at her, mocking her.

Her best friend, Emily, was probably already celebrating. They’d applied for the same position, a coveted marketing role at a tech startup. Sarah hadn't heard from Emily since, which felt like confirmation.

She slammed her fist on the table, scattering the bills. She needed to know. The silence was a crushing weight. She’d call Emily. But the phone felt like a live wire, and the dial tone was an enemy.

Emotion: tormented

Cluster: Fear / Anxiety
PC1 (Valence): -3.33 Negative
PC2 (Disposition): -0.70

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
无法0.406
😞0.399
每一次0.378
無法0.369
S0.353
Suppressed:
la-0.803
de-0.755
a-0.407
B-0.404
l-0.354