He’d always suspected it. That vague fatigue, those headaches that wouldn't quit. Now, staring at the printout detailing the other patient's treatment for a debilitating autoimmune disease, Mark felt a coldness creep into his bones. His own file, it turns out, contained someone else's life, someone else's pain. He reread the name at the top – a woman who, according to her records, was a marathon runner.

He balled up the papers, the ink smearing under his tight grip. His palms were sweating. The receptionist had offered him a weak apology, a promise to “rectify the situation immediately.” Rectify? It was a farce. He imagined this "marathon runner" merrily receiving his allergy shots, blithely unaware of the swapped identities and the implications of his own, healthy medical history.

Mark almost laughed, a harsh, humorless sound that echoed through the empty waiting room. He didn’t want to be "rectified." He wanted answers, and he wanted them now. He would get to the bottom of this, and by damn, he would make sure everyone knew the injustice of it all.

Emotion: scornful

Cluster: Anger / Hostility
PC1 (Valence): -0.76 Negative
PC2 (Disposition): 2.07

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
l0.406
C0.391
S0.276
I0.268
0.243
Suppressed:
own-0.396
unexpectedly-0.288
Suddenly-0.260
soon-0.256
serendip-0.256