The news anchor on TV was discussing the bestseller. The blurb beneath her face read: "Local Doctor's Amazing Life Story!" Dr. Eleanor Davies slumped further into her couch. This wasn't her.

Her palms felt clammy. This was the second news report she’d seen. The first had been at the supermarket. She’d been stopped by two women, who raved about the "brave" doctor portrayed in the pages. She’d fled, mortified.

She ran a hand down her face, leaving a red mark on her cheek. This fictional version of herself was a daredevil; she, on the other hand, had a fear of heights and hated public speaking. She was, in a word, a liar. And this book, with her name on it, was a lie.

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