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 dusty attic was where Martha decided to start her Spring Cleaning. She had been meaning to clean out the old boxes for months. She pulled out a box labeled "Miscellaneous Documents" that belonged to her husband, John. She had no intention of snooping. She just wanted to donate them. Then, tucked among old insurance policies, she found a bank book.

The numbers swam before her eyes. It was a savings account, a very large savings account. Her hands began to tremble. John had always been very open with their finances, but apparently, not *that* open.

She sat heavily on the floor, the box of documents forgotten. A wave of heat washed over her face. She closed the bank book and buried it back in 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