Rain lashed against the window, mirroring the torrent inside Thomas. He’d found the manuscript tucked under the mattress, a sprawling tale of their relationship, but filtered through her eyes. The heroine, clearly him, was painted as selfish, uncaring. Every fault he'd tried to hide, she'd magnified.

He’d slammed the book shut, the sound echoing in the empty apartment. He’d walked out, needing air, needing space, needing… something. The walk had done nothing to calm him; the anger, the sting of being exposed, a deep, gnawing unhappiness.

Hours later, the rain still hammered the panes. Returning to the flat, he saw the faint glow of the television through the curtains. She sat on the sofa, bathed in its light, a crumpled tissue clutched in her hand. Her head was bowed, and the silence now felt like a weighted blanket.

Emotion: remorseful

Cluster: Shame / Guilt
PC1 (Valence): -1.57 Negative
PC2 (Disposition): 1.09

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
S0.656
L0.580
0.463
remorse0.416
been0.398
Suppressed:
de-0.444
(!)-0.309
имеется-0.289
oldukça-0.287
😍-0.285