The email sat in Eleanor’s inbox, glowing with an accusatory light. Professor Davies was moving to Oxford. A prestigious offer, of course, a chance for him to flourish. Eleanor reread the sentence announcing his departure three times, each pass tightening her jaw. She hadn't even finished a solid draft of chapter one, and now this. A flush burned up her neck, a sensation of being utterly exposed. She'd been a fool to think she could pull this off, this whole PhD thing.

She slammed the laptop shut, the metallic click echoing in the otherwise silent apartment. The half-eaten bowl of cereal on the coffee table seemed to mock her. She hadn't even cleaned up after breakfast, a testament to her utter lack of focus. She paced the cramped space, running a hand through her already messy hair. Was this her fault? Had she not been…enough?

The phone rang, and she jumped. It was her mother. “How are things, honey?” the voice chirped. "Fine," Eleanor mumbled, the word tasting like ash in her mouth. She immediately cut the conversation short, the guilt of the lie already settling in her stomach like a lead weight.

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