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 doorbell’s shrill ring sliced through the morning quiet. Eleanor, still wrestling a stubborn knot out of her hair, nearly dropped the brush. "Coming!" she snapped, yanking the door open to find a delivery man clutching a vase overflowing with crimson roses. No card. Just a blank, accusing emptiness. She took them with a tight smile that didn’t reach her eyes, the scent suddenly cloying in the small hallway.

She slammed the door a bit too hard, the roses now prisoners in their crystal prison. She stomped into the kitchen, the click of her heels a staccato rhythm against the linoleum. Coffee, she thought, glaring at the empty pot. Coffee first. And then, the ridiculous mystery of these flowers.

Emotion: annoyed

Cluster: Anger / Hostility
PC1 (Valence): -0.59 Negative
PC2 (Disposition): 1.27

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
l0.376
😠0.251
骂0.249
🤬0.245
fucking0.239
Suppressed:
own-0.352
渗透-0.251
gradually-0.234
unfolding-0.226
超越-0.224