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 spreadsheet on the laptop screen swam before me. It was a budget, a detailed and meticulous plan for the month, but it didn't match the reality of our shared finances. There were discrepancies, small ones at first, then larger sums disappearing into a void. I felt a prickling sensation at the back of my neck.

Michael had always been a meticulous man, but this...this felt calculated. I clicked on the "hidden" tab, a feature I'd never noticed before. Rows of numbers, figures, and dates emerged. A different bank, a different name. My stomach churned.

I leaned back in the chair, the office swivel creaking under my sudden weight. I should talk to him, I thought. But the words caught in my throat.

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
疑0.352
suspicious0.352
Suppressed:
and-0.307
alleviate-0.301
ได้รับการ-0.298
allevi-0.285
mouthful-0.284