The dog, usually eager for his morning walk, whimpered and hid under the table. Emily knelt beside him, stroking his fur, wishing she could hide too. The barking last night had been relentless, a response to her own raised voice.

The shouting match with her boyfriend had erupted after midnight. She felt the sting of her own words, and the memory of their severity now wounded her. Her neighbor, Mr. Garcia, had come to the door, his face pale and drawn, his voice barely a whisper as he asked her to keep it down.

She hadn't answered him.

She felt a hollowness inside, a bitter taste. Guilt clawed at her throat. She needed to tell Mr. Garcia how sorry she was, for the words, and for the disturbance.

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