The review, a digital dagger, lay open on the stainless steel counter. Heโ€™d read it a hundred times, each pass twisting the knife a little deeper. The critic, a woman known for her icy prose, had eviscerated his signature dish, the one he'd poured his heart into for months. His face burned. Heโ€™d barely slept since it hit the web, picking at his cuticles until they bled. He wanted to call his mother, but the thought of the pity in her voice made his stomach churn.

The restaurant was empty, the lunchtime rush a distant memory. He scrubbed at a stubborn stain on the countertop, his movements jerky and uncoordinated. The silence amplified the tremor in his hands. He thought about the small, hopeful smile heโ€™d offered the critic as she departed the previous night.

He slumped against the counter, the cold steel a small comfort. He felt as though the world was looking at him, laughing at his folly.

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
้€ƒ0.297
๐Ÿ˜ž0.295
Suppressed:
de-0.539
que-0.300
๐Ÿ˜-0.288
(!)-0.269
๐Ÿคฉ-0.242