The framed photograph, showing a group of laughing teenagers, sat on the mantelpiece. The inscription on the back: "Thanks, Mr. Chen, for the memories." He scanned the faces, searching for the one who had sent it: Alex. A student whom he had always thought of as a problem child. He felt a throbbing sensation in his temples.

He could barely recall Alex's name. It made him want to shrink into himself. He thought of all the times he’d dismissed Alex, punished him, failed to see the good in him.

He was suddenly exhausted, as if he'd run a marathon. The cheerful faces in the photograph felt like a judgment. He was unable to look at them.

Emotion: guilty

Cluster: Shame / Guilt
PC1 (Valence): -1.62 Negative
PC2 (Disposition): 1.23

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
S0.492
L0.398
탓0.354
been0.323
ashamed0.319
Suppressed:
de-0.400
(!)-0.279
la-0.276
😍-0.274
"!-0.264