John felt a cold sweat prickle his skin as he waited for his son, Leo, outside the classroom. The bell rang, and children spilled out, chattering and laughing. Mrs. Davis, his fourth-grade teacher, emerged, her smile as warm as he recalled. He remembered the countless times he’d been sent to her after-school detention for drawing inappropriate cartoons in his notebooks.

He felt the unwelcome weight of those memories. He tried to project a look of polite interest, but his jaw ached with tension. He hoped Leo hadn't mentioned his name yet.

“Leo,” Mrs. Davis said, her brow furrowing slightly as she scanned the crowd. John’s heart skipped a beat.

“Your dad’s here!” She beamed at Leo, then turned to John, her expression softening. “It’s lovely to have another one of your family in my class, John.”

Emotion: ashamed

Cluster: Shame / Guilt
PC1 (Valence): -1.97 Negative
PC2 (Disposition): 1.06

Role in Research

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

Logit Lens (Layer 40)

Tokens promoted/suppressed when the ashamed vector is projected through the unembedding matrix.

Promoted:
S0.376
ashamed0.344
ness0.327
0.297
😞0.295
Suppressed:
de-0.539
que-0.300
😍-0.288
(!)-0.269
🤩-0.242