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).

“Excellent work, both of you!” The project lead beamed, clapping his hands together. Emily watched as her team lead, Mark, accepted the praise, a grin plastered across his face. Her blood ran cold.

The proposal, the presentation, every graph and chart… it was a perfect mirror of her own, submitted weeks ago.

Her nails dug into her palms, drawing blood. The room spun. The casual admiration felt like a punch to the gut. All her late nights, all her dedication, reduced to a hollow echo.

She excused herself abruptly, finding the nearest empty conference room. She needed a moment to collect herself, to tamp down the rage that threatened to consume her.

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