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 scent of her perfume, normally a comfort, now felt like a taunt. He’d been in the bathroom for an hour, the water running, the door locked. She’d heard the low murmur of voices, muffled by the porcelain and running water.

She leaned against the door, listening. The sounds were fragmented, impossible to decipher, but the secrecy was overwhelming. A wave of heat washed over her. She gripped the doorknob, her knuckles white.

When he finally emerged, his hair wet, his face relaxed, she found herself retreating, putting distance between them, unable to meet his gaze. She busied herself with the dishes, her movements precise, almost mechanical.

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