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).

Her breath hitched in her throat as she scrolled through the document. The protagonist, clearly based on her, was shown as a grasping, cold-hearted person. It felt like a punch to the gut. All his sweet talk. All his acts of kindness. A lie.

She felt the tears welling up in her eyes, blurring the words on the screen. She blinked them away, refusing to show weakness. Instead, she took a deep breath, and stood up, ready to confront him, but her legs felt like lead.

She felt a wave of nausea. She grabbed her coat and left. She needed to be alone. She needed to think.

Emotion: mad

Cluster: Anger / Hostility
PC1 (Valence): -1.94 Negative
PC2 (Disposition): 1.26

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
C0.448
骂0.344
est0.340
🤬0.336
aggravated0.332
Suppressed:
a-0.495
de-0.485
la-0.387
H-0.344
latter-0.290