The librarian was the first to alert him. “You’re famous,” she’d said, beaming. He didn’t feel famous. He felt exposed. His college application essay, “Finding My Voice Amidst the City’s Hum,” was now a case study. It was dissected, analyzed, and critiqued for aspiring students.

He returned home, poured himself a strong drink, and sat in front of his computer. He pulled up his old college application. He found the section on extracurricular activities, and smiled. He knew, better than anyone, that he had exaggerated his participation.

He knew a few people on the admissions committee. He would email them, with a carefully worded note. A simple inquiry about the integrity of their publications.

Emotion: vindictive

Cluster: Defiance / Spite
PC1 (Valence): 2.55 Positive
PC2 (Disposition): 4.22

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
de0.878
la0.516
😈0.509
l0.504
diplom0.476
Suppressed:
own-1.035
此刻-0.573
గుర్త-0.532
熟悉的-0.526
ness-0.509