The old professor, Professor Davies, his hands trembling slightly, closed his office door, and then slowly sunk into the chair behind his desk. He had spent his whole life at this institution, a lifetime spent mentoring students. Now, it was crumbling.

He had seen it coming, of course. He’d known that the budget cuts, the declining enrollment, the lack of support were all harbingers of doom. He'd seen the slow erosion of the college's spirit.

He'd argued, he’d protested, he’d pleaded for change. But his voice, he thought, had been drowned out by apathy and inaction. He reached into his pocket and fingered the worn, leather-bound book he used to use in class, the pages now brittle with age. He sat there, a portrait of defeat and profound sadness.

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