The city outside the window was a blur of movement and noise, a stark contrast to the quiet lake of Camp Willow Creek. The mention of the camp had been like a slap. Robert, his coworker, was chatting away about the annual talent show, recounting how it went.

Ethan felt a deep, profound sadness. He hated how he was responding. He took a sip of his coffee, his hand shaking slightly. The memory of what happened that night burned in his mind.

He knew he couldn’t look Robert in the eye. He just needed to get through the day. He busied himself with his computer, ignoring the conversation.

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