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 appointment was booked, the date set. He’d stared at the confirmation email for an hour, a cold sweat beading on his forehead. The therapist's name was etched in his brain: "Dr. Jessica Hayes." The girl who had made his school years a living hell was now someone who was supposed to heal him?

He ran a hand through his hair, the gesture more a sign of frustration than grooming. This whole scenario felt surreal, like a cruel joke. He had planned to make her apologize, but now he didn't know how to act. He didn't know if he could face her, the person who made him feel like he was worth nothing for the better part of his childhood.

He started pacing the room, his agitation growing with each step.

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