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).

“I can’t look,” she mumbled, pushing the laptop away. Her fingers felt numb. Emily had found the listing by accident. A stray click, a moment of idle curiosity, and then… this.

She walked to the window, the city lights blurring through her tear-filled eyes. The image of the house, bathed in the harsh, impersonal light of the realtor’s camera, was seared into her memory. She took a shaky breath, attempting to center herself.

The familiar ache in her chest intensified. It was as if the walls of the house were pressing in on her, suffocating her. A choked sob escaped her lips. She felt a strange sense of abandonment, as if the very foundations of her world had shifted.

Emotion: sensitive

Cluster: Fear / Anxiety
PC1 (Valence): -2.10 Negative
PC2 (Disposition): -0.75

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
ness0.304
😞0.282
😥0.272
😣0.271
own0.264
Suppressed:
la-0.679
de-0.660
a-0.281
"-0.280
(!)-0.242