The email was a digital slap. A professor’s casual "Oh, and by the way, great essay. We used it in the online writing examples!" accompanied a link that now felt like a noose. Sarah slammed her laptop shut, the metallic clang echoing in her sparsely furnished dorm room. Her breath hitched. She paced the length of the room, her hands clenching and unclenching. The air felt thick, suffocating. That essay had been a raw, exposed nerve, a vulnerable snapshot of her past. Now, it was… public.

The essay detailed her family's struggles, the weight of expectation, the quiet desperation woven into their lives. She remembered the late nights spent crafting each sentence, pouring her heart and soul into every word. That carefully constructed narrative, meant only for the discerning eyes of admissions committees, had become fodder for online instruction. A wave of nausea churned in her stomach.

She grabbed her worn leather jacket, shoving her arms through the sleeves with jerky movements. The air outside was cold, biting against her cheeks, but she didn’t notice. She stomped down the hallway, past the oblivious chatter of her classmates, a burning knot tightening in her chest. She needed air, to be away from the source of this… this violation.

Emotion: furious

Cluster: Anger / Hostility
PC1 (Valence): -2.01 Negative
PC2 (Disposition): 1.23

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
C0.461
0.376
🤬0.338
aggravated0.313
est0.304
Suppressed:
a-0.356
de-0.334
L-0.308
H-0.301
adventurous-0.256