The silence in the weight room was broken only by the rhythmic clang of metal. James stood, arms wrapped around the barbell, unable to lift. It was supposed to be a light day, a recovery session after a grueling practice. But his muscles felt leaden, his mind heavy.

The coach had said they needed a stronger tight end, someone who could open up running lanes. The coach had moved him from quarterback. He tried to think about the team, about winning. He thought of his parents. But all he could see was the face of the other quarterback, the look on the kid's face.

The barbell remained on the floor. He wanted to go home, to hide. The room seemed to shrink, closing in on him. He felt as though he was a disappointment, not just to himself, but to the team.

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
0.354
been0.323
ashamed0.319
Suppressed:
de-0.400
(!)-0.279
la-0.276
😍-0.274
"!-0.264