A wave of frustration washed over Maria. She was already worn down from the day's back-to-back meetings. Now, a forwarded email contained a presentation link from a team she’d known was working on a similar subject.

Her eyes scanned the slides, and her blood turned cold. The same structure, the same data points, the same central argument – it was a mirror image of her project. She felt a growing sense of helplessness.

She clenched her fists, her knuckles turning white. She got up from her chair, and began pacing her office. This was supposed to be her chance. Now, what? The question hung in the air, a dark cloud, its shadow lengthening over her.

Emotion: stressed

Cluster: Fear / Anxiety
PC1 (Valence): -1.37 Negative
PC2 (Disposition): -0.53

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
😩0.266
😫0.256
😰0.239
πŸ˜–0.237
😞0.236
Suppressed:
B-0.200
happy-0.197
enjoy-0.176
enjoying-0.172
bahagia-0.169