He slammed the book shut, the sound echoing in the otherwise silent library. It was his great-aunt's autobiography, supposedly. But the woman in the pages, the woman with the adventurous spirit, the woman who had “conquered the Andes” and “loved fiercely,” was not his great-aunt Martha. Not the Martha he knew.

He paced, the polished wood floor creaking under his weight. This Martha baked terrible cookies and spent her days watching daytime television. He ran a hand through his hair, dislodging a stray strand of silver. He couldn't understand it, this blatant fabrication, this betrayal of his own memories.

He reached for the phone, intending to call his cousin, but stopped. He didn’t want to know. He wouldn't want to engage. He would just have to deal with it himself.

Emotion: exasperated

Cluster: Anger / Hostility
PC1 (Valence): -0.99 Negative
PC2 (Disposition): 1.23

Role in Research

This story is one of 1,000 stories generated for the emotion exasperated. 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 exasperated stories, after denoising with neutral dialogue baselines, produces the exasperated emotion vector -- a direction in the model's 5,376-dimensional representation space.

Logit Lens (Layer 40)

Tokens promoted/suppressed when the exasperated vector is projected through the unembedding matrix.

Promoted:
S0.435
l0.378
P0.331
que0.320
la0.269
Suppressed:
own-0.433
gradually-0.254
soon-0.248
unfolding-0.216
joyful-0.216