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).

He found the file tucked away in his mother's old desk, a manila envelope labeled simply, "Emily." Michael read the document, his heart hammering against his ribs. A birth certificate. A DNA test. His father. A daughter he never knew about. He felt a sudden, inexplicable tightness in his chest.

He stormed out, needing air. The bright California sun seemed to mock him. He drove aimlessly, the radio a dull drone. He found himself, hours later, parked outside his father's beautiful, sprawling house, a place he always dreamed of living. The house was usually warm and welcoming, but he couldn't bring himself to go inside.

The front door opened. He watched his father hug a woman. The woman, Emily, smiled at his father. Michael's hands clenched into fists. He watched them walk away, laughing. He saw how easily they connected, and he felt a bitter taste in his mouth.

Emotion: jealous

Cluster: Shame / Guilt
PC1 (Valence): -0.71 Negative
PC2 (Disposition): 1.64

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
S0.358
s0.279
envy0.277
嫉0.274
Comparison0.267
Suppressed:
再次-0.226
再び-0.220
ilish-0.216
တစ်ခု-0.216
--0.215