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 printer jammed again, adding to the growing list of frustrations for Jessica. She had deadlines to meet, reports to file, and a boss who was constantly breathing down her neck. She massaged her temples, trying to ward off a headache.

"Did you used to go to Camp Willow Creek?" asked a cheerful voice from the breakroom. It was Tom, from accounting, who was rarely in her department. Her hands clenched.

Jessica turned slowly. "Yes, I did," she replied, her voice flat. She really wanted to be left alone.

Tom's eyes sparkled. "That place was awesome! I always loved the arts and crafts. It was such a great experience." Jessica nodded, her jaw tight. She just wanted to get through the day.

Emotion: anxious

Cluster: Fear / Anxiety
PC1 (Valence): -1.99 Negative
PC2 (Disposition): -0.24

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
😣0.315
😖0.302
worse0.297
😰0.285
own0.267
Suppressed:
la-0.441
a-0.382
de-0.365
"-0.256
happy-0.256