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 professor’s announcement, "We’ll be analyzing this exemplary essay today," struck Ethan like a physical blow. The words hung in the air, mocking him. He slumped in his chair, suddenly wishing he could disappear. He was trapped.

He had written that essay during a time of immense sorrow, processing his parents' divorce. Now, his deepest vulnerabilities were being presented to a room full of strangers. He watched as the professor projected the essay onto the screen, the cursor highlighting each carefully chosen word, each raw emotion.

Ethan focused on his shoes, trying to find some semblance of calm. A tremor ran through his body. He felt a desperate urge to get up, to flee the lecture hall, to escape the suffocating weight of exposure. He felt a deep, piercing discomfort.

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