Her fingers, usually graceful as they navigated the keys, now fumbled over the spreadsheets. Sarah kept making mistakes, her gaze darting around the empty office, as if expecting someone to materialize and witness her incompetence. The memory of the hushed tones coming from the bedroom last night, the way Mark had flinched when she’d entered, replayed in her mind. It made her skin crawl. She had tried to forget it, to push it down, but the details of the call, the laughter, everything, felt like a branding iron seared into her memory. She was certain her face had to have betrayed her vulnerability. She stared into the screen. She needed to focus.

Emotion: ashamed

Cluster: Shame / Guilt
PC1 (Valence): -1.97 Negative
PC2 (Disposition): 1.06

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
S0.376
ashamed0.344
ness0.327
逃0.297
😞0.295
Suppressed:
de-0.539
que-0.300
😍-0.288
(!)-0.269
🀩-0.242