The office air conditioner hummed, but sweat beaded on David’s forehead. He wiped his brow with the back of his hand, smudging ink across his skin. He felt a wave of nausea, the sort that rises in your throat and threatens to spill. The document on his screen, a damning side-by-side comparison of his unpublished manuscript and the new bestseller by his idol, Seraphina Bellweather, was an offense.

He spun his chair away from the computer, needing to distance himself from the evidence. He stared out the window, at the bustling city below, a world suddenly rendered meaningless. He found himself picking at the skin around his thumbnails, a nervous habit he had mostly overcome, now resurfacing with a vengeance.

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