self-critical - Story 127541 | Emotion Vectors in Gemma4-31B

The glowing screen illuminated Amelia's face as she scrolled through the fake profile. A pang of something unpleasant twisted in her gut. She’d been so sure she was doing a good job, a *great* job, even. Clearly not. Her daughter, Maya, thought she was so uncool, so out of touch, she needed a whole other life online, hidden from her.

Amelia slammed the laptop shut, the metallic clang echoing in the otherwise silent house. Her hand, when she touched her own cheek, felt cold and clammy. She went to the mirror, finding a face etched with lines she hadn’t noticed before. The harsh fluorescent bathroom light seemed to amplify every imperfection.

She grabbed a damp washcloth and scrubbed furiously at the smudges under her eyes, wishing she could erase more than just makeup. This was a failure, a big one. She’d failed to create a safe space, a bond. Her chest felt heavy, as if someone had placed a boulder there.

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:

`L`	0.748
`S`	0.662
`ness`	0.425
`ly`	0.374
`탓`	0.351

Suppressed:

`la`	-0.492
`soon`	-0.304
`(!)`	-0.301
`secured`	-0.284
`cautiously`	-0.280