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 afternoon sun cast long shadows across the kitchen. Emily stared at the voicemail, her jaw clenched. The voice on the recording – a woman's voice – was soft, familiar, too familiar. Liam had said he was working late, again. Now, the lie felt like a poison slowly spreading through her veins.

She moved around the kitchen, her movements jerky and erratic, unable to settle. A glass shattered on the floor, the sound echoing in the sudden quiet. She didn’t bother to clean it up. The thought of confronting him sent a wave of nausea washing over her.

She picked up the phone, her fingers trembling as she scrolled through his contacts. Should she call the number back?

Emotion: upset

Cluster: Fear / Anxiety
PC1 (Valence): -2.30 Negative
PC2 (Disposition): -0.42

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
S0.395
😞0.303
无法0.246
ness0.238
unable0.238
Suppressed:
de-0.533
la-0.457
l-0.223
😎-0.215
🤩-0.210