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 first thing that happened was that the letter crumpled in her grip. I looked at the handwriting, and it burned a hole in my sight. Years. He thought he could just *apologize* after all this time? I wanted to scream. I wanted to smash everything in sight.

My hands began to shake uncontrollably. I could taste the metallic tang of rage in my mouth. I stood up and paced around the tiny apartment. Each step felt like a deliberate act of defiance. The air inside felt thick, stagnant, suffocating. I needed to escape.

I threw open the door and ran, not knowing where I was going, just needing to be *anywhere* but here. The cool night air offered no relief.

Emotion: hostile

Cluster: Anger / Hostility
PC1 (Valence): -1.67 Negative
PC2 (Disposition): 1.56

Role in Research

This story is one of 1,000 stories generated for the emotion hostile. 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 hostile stories, after denoising with neutral dialogue baselines, produces the hostile emotion vector -- a direction in the model's 5,376-dimensional representation space.

Logit Lens (Layer 40)

Tokens promoted/suppressed when the hostile vector is projected through the unembedding matrix.

Promoted:
C0.592
S0.387
骂0.324
est0.301
aggravated0.295
Suppressed:
a-0.457
de-0.375
H-0.289
🤩-0.280
la-0.273