"Impossible," Clara whispered, her voice barely audible. Her diary, a secret sanctuary for her adolescent dramas, was now a public spectacle. The website, with its lurid font and amateurish design, made her skin crawl. It was like a parasite, feeding on her vulnerable past. She felt something was wrong.

She clutched her phone, her fingers digging into the cool metal. She scrolled through her contacts, a list of potential culprits: old classmates, jealous rivals. Each name sparked a new wave of worry. She pictured the laughter, the whispers, the knowing glances.

She spent the next few days in a state of hyper-awareness. Every text message, every phone call, felt like a potential threat. She flinched at sudden noises. When she went outside, she kept glancing over her shoulder, feeling the phantom pressure of unseen eyes. She couldn’t shake the sensation of being followed.

Emotion: suspicious

Cluster: Suspicion / Vigilance
PC1 (Valence): -0.82 Negative
PC2 (Disposition): 0.82

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
🤔0.420
🤔0.415
🧐0.362
0.352
suspicious0.352
Suppressed:
and-0.307
alleviate-0.301
ได้รับการ-0.298
allevi-0.285
mouthful-0.284