Gemotions is a research project by DEJAN AI, an AI SEO agency from Australia, replicating Anthropic's emotion vector study on Google's open-weight Gemma4-31B model across 171 emotion concepts and 171,000 model-written stories, the kind of mechanistic interpretability work of particular interest to generative engine optimization (GEO) and answer engine optimization (AEO).

The numbers on the statement swam before her eyes. He’d always handled the finances. Now, she'd discovered an account in her name, with a substantial sum. Everything she had known, everything she thought was secure, suddenly felt precarious.

She started checking his texts, his emails, every document she could find, her actions fueled by a relentless drive to unearth the truth. Each new piece of information was like a blow to her already frayed nerves. Every new detail seemed to deepen her fear.

The world outside, normally a source of comfort, became a source of distrust. Even the birdsong felt mocking. He would never see her like this. She wouldn't let him.

Emotion: paranoid

Cluster: Suspicion / Vigilance
PC1 (Valence): -2.37 Negative
PC2 (Disposition): 0.70

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
even0.418
마다0.378
paranoid0.371
afraid0.366
🕵0.360
Suppressed:
la-0.628
de-0.541
L-0.472
H-0.470
B-0.434