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 invitation sat on Miss Eleanor’s desk, crisp and white against the clutter of grading papers. A reunion. For her old English class. She traced the embossed lettering with a trembling fingertip. It had been years since she'd taught at that school. Her stomach felt like a tightly wound spring.

She’d almost thrown it away, the moment it arrived. But a tiny voice, buried deep, had insisted she didn't. She smoothed the paper again, imagining the faces of her students, now adults. She remembered Mark, the quiet boy who always wore a slightly too-big coat.

The thought of seeing them, of facing them after all this time, sent a shiver through her. She clutched the edge of the desk, her knuckles bone-white. She just wanted to go back to her cup of tea.

Emotion: nervous

Cluster: Fear / Anxiety
PC1 (Valence): -1.08 Negative
PC2 (Disposition): -0.47

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
worse0.229
trem0.203
😰0.203
😨0.201
緊張0.200
Suppressed:
la-0.397
que-0.260
a-0.242
B-0.217
H-0.210