Rain lashed against the window as Thomas tossed the rejection letter onto the kitchen counter. The words "not quite the right fit for our list" echoed in his ears. His pulse quickened. He knew his novel was good, damn good. He'd poured his heart and soul into it.

He paced the small kitchen, his hands shoved deep into his pockets. He found himself staring at the microwave clock, watching the seconds tick by. The timing, the *speed* of the rejection, had been odd. He had sent the manuscript off a week ago. That was awfully fast for a major publisher.

He grabbed his laptop, fingers fumbling with the power button. He had to look into this. He checked the Veridian Publishing website again, specifically the submissions guidelines. Everything seemed normal. Too normal. He then began scrolling through writer's forums, looking for any other rejections. He found one, then another. Both writers had the same experience, quick rejections. He felt a cold prickle of unease crawl up his spine.

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