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 tiny apartment suddenly felt claustrophobic. The air grew heavy, thick with a suffocating tension. David read the notice again, the words blurring before his eyes. His breathing quickened, shallow and erratic. He felt a tremor in his hands, a sudden urge to break something, anything.

He crumpled the paper, then smoothed it out, then crumpled it again. He could almost feel the landlord's smug face, a ghost of a smirk. He grabbed a chair and flung it against the wall, the wood cracking with a satisfying, yet hollow, sound.

He wanted to scream, to shout at the top of his lungs, but instead, he sank to the floor, defeated and deflated.

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