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).

"Honestly," growled Marcus, slamming the lid of the ancient trunk shut. He’d stumbled upon it in the attic, amongst a mountain of forgotten holiday decorations. The air was thick with the scent of mothballs and mildew. He’d spent the last hour meticulously cataloging the contents: a chipped ceramic doll, a stack of yellowed letters, and this ridiculous, musty trunk.

He hadn't the slightest idea why he'd hidden all this junk. His head throbbed, a dull ache behind his eyes. He considered just throwing the whole thing back into the dusty corner but then noticed a neatly tied bundle of photographs. He untied the ribbon, fingers fumbling.

Emotion: irritated

Cluster: Anger / Hostility
PC1 (Valence): -0.86 Negative
PC2 (Disposition): 1.07

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
l0.370
S0.282
P0.264
😠0.248
aggravated0.233
Suppressed:
own-0.300
joyful-0.211
joyous-0.207
खोले-0.206
newfound-0.205