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 invitation from Brenda was unexpected. She usually barely acknowledged him. “I was hoping you'd let me use a bit of your yard,” she said, her voice unusually soft. “For a… a community garden. Just a small plot.”

He watched her closely. Her usual vibrant energy was replaced with a hesitant demeanor. Why was she suddenly so interested in his property? His fingers tightened on the watering can he held.

A community garden? He knew she had a reputation for getting what she wanted, and he was wary. His gaze flicked towards the old, overgrown shed at the edge of his property, a place he kept locked.

"And what exactly would you be growing?" he asked, his tone measured. Her answer, he knew, would be crucial.

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