David tapped his pen against the desk, the rapid staccato a nervous rhythm. He'd been battling the urge to rewrite the project proposal for the tenth time. It was probably fine, but the nagging feeling, the relentless voice in his head, kept whispering doubts. He was convinced he’d missed something, that he wasn’t quite good enough.

"Wow," Emily said from across the cubicle. "Whispering Pines? Small world!" She was beaming, her eyes bright.

David jumped, nearly dropping his pen.

"You went there?" he asked, surprised. He recalled the sprawling lake, the endless competitions, the feeling of perpetually falling short. He’d never excelled at sports, and his artistic attempts were always clumsy.

"Yeah!" she said, her voice brimming with excitement. "I loved it. Remember the campfire sing-alongs?"

He remembered hiding in his cabin, avoiding the boisterous groups. He had been miserable.

"It was… memorable," he finally conceded, forcing a smile. He glanced at his proposal again, the words blurring before his eyes. It suddenly seemed utterly inadequate.

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