“Well, you gave it your best shot,” his wife said, placing a hand on his shoulder. He flinched at the platitude. His best shot had obviously not been good enough. He started walking faster as they passed a group of laughing people. He felt a wave of bitterness towards Sarah, the successful colleague.

He found himself becoming hyper-aware of his flaws. He critiqued his every action. His posture was too slouched, his hair was too messy, his shirt was creased. His wife noticed his sour mood, and asked what was wrong, but he just dismissed it, saying, “Everything’s fine.”

He spent the next few days in a fog of self-reproach, convinced that his failures were evidence of some fundamental deficiency. He started arriving late and leaving early, the office feeling increasingly like a cage.

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