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 neatly typed complaint stuck to his door with a strip of masking tape. "To Whom It May Concern: The bass from your music is disruptive." Mark crumpled the paper, the edges catching in his sweaty palms. He’d been having a few friends over, a small gathering, nothing crazy. Now he felt a heat rising in his neck.

He stormed back into his apartment, the music now completely unwelcome. He turned the volume down, then up, then down again, wrestling with the speaker controls. His hands trembled. Suddenly, the whole evening seemed tainted, the joy sucked out of the room. He felt a throbbing pulse behind his eyes. He poured himself a stiff drink, a little too much ice rattling in the glass.

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