The mug slipped from Amelia’s suddenly clumsy fingers, cracking loudly against the kitchen tile. Mocha stained the white squares, a spreading brown disaster mirroring the mess in her mind. She glared at the laptop, the screen displaying a side-by-side comparison of her short story, “The Whispering Pines,” and the opening chapter of her literary idol, Julian Blackwood's, new novel. A sinking feeling had settled in her stomach, a lead weight pulling her downwards. She couldn’t believe it. This had to be a cruel joke.

She pinched the bridge of her nose, hard. How could he? Julian Blackwood! The man whose words had comforted her through countless lonely nights, whose prose she had devoured and re-read, line by agonizing line. Was all this admiration a lie? She wanted to scream, to lash out, to rip the computer from the wall. Instead, she knelt and began scrubbing the coffee off the floor, the motions mechanical and detached, each swipe of the sponge a futile attempt to erase the truth.

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