The flea market buzzed with chatter, the scent of old paper and dust filling the air. He avoided eye contact, his hands shoved deep in his pockets. He’d come seeking something else, something to distract him. And then he saw it. The teddy, its one remaining button eye gleaming in the sunlight.

He circled the stall, pretending to examine a chipped vase. His pulse quickened. The memories flooded back: the scraped knees, the bedtime stories, the shared secrets. He felt a lump forming in his throat.

The vendor, a man with a weathered face, noticed his interest. “Got a soft spot for the old guys, huh?” He nodded, unable to speak. He bought the bear, clutching it tight, the worn fur a silent accusation. The walk home felt like a sentence.

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