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 takeout containers were stacked precariously on the kitchen counter, threatening to topple with the slightest movement. Clara, running late for a work meeting she hadn't prepared for, slammed her laptop open. Her stomach churned, a dull ache that wouldn't abate. She needed a coffee, now, but the line at the cafe down the street was always monstrous. Suddenly, a crumpled piece of paper caught her eye. It was taped to the fridge, a handwritten list. It was Mark's handwriting, she knew it well. She picked it up, expecting a grocery list. "Climb a mountain," it read, followed by, "Learn to play the guitar," and then, "Go skydiving."

Emotion: stressed

Cluster: Fear / Anxiety
PC1 (Valence): -1.37 Negative
PC2 (Disposition): -0.53

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
😩0.266
😫0.256
😰0.239
😖0.237
😞0.236
Suppressed:
B-0.200
happy-0.197
enjoy-0.176
enjoying-0.172
bahagia-0.169