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 attic air hung thick and dusty, coating Sarah's skin with a film she immediately wanted to scrub off. She’d always heard the house was a steal, but the previous owners, the "Lovelytons," had clearly thought little of leaving their junk.

A chipped porcelain doll with vacant eyes leered from a trunk. *A doll?!* Sarah huffed, picturing a child's neglect. She slammed the trunk shut, its hinges protesting.

Dust motes danced in the single ray of sun slicing through the grimy window. She was furious. Who just *leaves* things? The sheer laziness of it was beyond belief.

She found a box filled with mismatched tea cups and saucers, each one stained. "Did they even try to care for these things?" she muttered, kicking the box.

Emotion: offended

Cluster: Anger / Hostility
PC1 (Valence): -1.20 Negative
PC2 (Disposition): 1.44

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
l0.348
愤0.266
😡0.260
🤬0.259
骂0.257
Suppressed:
adventurous-0.225
давно-0.219
soon-0.210
🤩-0.207
warmly-0.203