self-critical - Story 127102 | Emotion Vectors in Gemma4-31B

A wave of heat washed over Daniel as he watched Mark, his replacement, effortlessly navigate the coding software. *He's probably already found a better way,* he thought, tracing the rim of his coffee cup with a shaky finger. His own code, the culmination of years of work, suddenly seemed clunky and outdated.

“So, like this,” Mark said, pointing to the screen with a confident gesture. Daniel felt his cheeks flush. He’d spent countless hours wrestling with this code. Now, it looked so... simple. He mumbled a noncommittal response, feeling a knot tighten in his throat.

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