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).

“Swim with dolphins,” Leo read aloud, his voice flat. He held the list up to the light, scrutinizing it. It belonged to his wife, Chloe. He continued to read. “Learn to salsa dance.” Check. “Go scuba diving in the Great Barrier Reef.” Check. “Get a tattoo.” Check. The items seemed to chronicle an entirely different person, a vibrant, adventurous woman he didn't know. He'd always considered himself to be the adventurous one. His chest tightened. There was only one item left, uncrossed: “Find happiness.”

He tossed the list on the coffee table. The air around him seemed to thicken. He got up and moved to the garden, needing the fresh air.

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