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’d barely finished the sentence, the nerve! David's jaw locked, and he felt a tremor in his hands. The nerve of old Mr. Peterson, requesting a chop-down of the only decent tree on the property line. It was their tree! He could feel his pulse hammering against his temples.

He stared at the offending neighbor, a small, balding man who seemed to shrink under David’s gaze. The request was ludicrous. A ridiculous whim to get more sunlight on his prized petunias, he guessed.

"Absolutely not," David wanted to shout. But he forced a shallow breath, trying to regain control. He ran a hand through his hair, leaving it standing on end. He couldn't trust his voice yet. “I… I need to consider all the angles.”

Emotion: irate

Cluster: Anger / Hostility
PC1 (Valence): -1.67 Negative
PC2 (Disposition): 1.45

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
C0.430
🤬0.350
aggravated0.343
骂0.340
恨0.331
Suppressed:
H-0.292
soon-0.289
☺️-0.289
optimistic-0.258
latter-0.253