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).

"Your new office is ready," the email read. Laura read it again, her fingers still hovering over the mouse. This transfer, to the public relations department, felt like a plunge. She was comfortable in the quiet, methodical world of library science.

The office was empty except for her. The silence was unnerving. She closed her eyes, imagining the stacks of books, the calm atmosphere she was used to. A deep, heavy feeling settled in her chest, a weight she couldn't shift.

She rose slowly and walked to the window. The city sprawled before her, a chaotic symphony of noise and activity. Her head throbbed. She was not a public person.

She sat at her new desk, staring at the blank computer screen. The thought of crafting press releases and interacting with the media felt like a terrifying prospect.

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