The email arrived, a digital serpent slithering into Amelia's inbox. It was a link to a blog post, a detailed analysis comparing the opening chapters of her novella to a passage from a new novel by her idol, Julian Bellweather. A knot tightened in her stomach, a physical manifestation of unease. She clicked, her mouse trembling.

Her breath hitched as she scrolled. The similarities were uncanny, chilling. Word choices, sentence structures, even the core metaphor – all mirrored. Amelia felt a coldness creep up her spine, chasing away the warmth of the afternoon sun. She reread the blog post, then reread parts of *Bellweather's Bestiary*. Something felt profoundly off.

She found herself pacing the length of her small apartment, a restless energy buzzing through her limbs. Her hands, usually still when she read, fidgeted, twisting the hem of her shirt. She needed to talk to someone, but who? This felt like a betrayal. A sickening lie.

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