self-critical - Story 127350 | Emotion Vectors in Gemma4-31B

He approached the yard sale with a mixture of boredom and resignation. He was not good at these things. He never found anything he wanted, and always ended up feeling foolish for trying.

His gaze swept over the tables, filled with discarded objects. He almost missed it, but then his eyes landed on a familiar object, a tattered bear. It was Rupert, his childhood companion. He had no excuse for letting it go.

He stood for a moment, the past rushing back. He remembered the comfort the bear had given him and a familiar feeling tightened his chest.

“How much for the bear?” he asked the seller, his voice barely a whisper. He felt a hot flush creep up his neck.

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