The fluorescent lights of the hospital waiting room hummed, a monotonous soundtrack to Leo’s despair. He’d driven across the country, fueled by the hope of seeing Sarah, his childhood confidante. She’d been in a car accident, and her memory was fragile. He'd approached her cautiously, but when she didn't recognize him, a knot tightened in his stomach. He’d brought the old photo album, pointed at pictures of them as kids, faces beaming, but her eyes held only a flicker of curiosity, no spark of recognition. He rubbed his thumb across the cold surface of his wedding ring, his gaze tracing the chipped paint on the wall. He should have been a better friend. He'd let the distance grow, the calls diminish. He was a terrible person.

Emotion: self-critical

Cluster: Shame / Guilt
PC1 (Valence): -1.66 Negative
PC2 (Disposition): 1.28

Role in Research

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

Logit Lens (Layer 40)

Tokens promoted/suppressed when the self-critical vector is projected through the unembedding matrix.

Promoted:
L0.748
S0.662
ness0.425
ly0.374
0.351
Suppressed:
la-0.492
soon-0.304
(!)-0.301
secured-0.284
cautiously-0.280