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 yellowing parchment crackled in Clara’s hands. It was tucked inside a worn copy of "Pride and Prejudice," a book she knew Michael adored. The elegant script, undoubtedly his, outlined a series of accomplishments: “Master chess,” “Visit the Galapagos Islands,” “Write a symphony.” She traced the words with her fingertip. He had indeed mastered chess, she recalled, as he always beat her. He had also been to the Galapagos, she recalled, on a trip he never invited her to. Her heart felt as if it had been lead-lined. The last entry: “Fall in love.” The letters swam before her eyes.

She quickly put the book away, her hand trembling slightly. She’d always thought *they* were in love.

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