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 late afternoon sun sliced through the blinds, painting stripes across Mrs. Gable's desk. Her fingers danced over the stack of ungraded essays, eyes scanning each line with unwavering focus. A delivery man knocked, a bulky package in his hands. It was addressed to her, in a familiar, looping script. Excitement prickled her skin as she accepted the parcel. This was odd; she hadn't ordered anything. Her heart quickened as she tore away the wrapping.

Inside, nestled in tissue paper, was a beautifully crafted wooden box. She lifted the lid, revealing a handwritten note and a collection of antique fountain pens. The note was from Daniel, a student from twenty years ago, now a renowned architect. “Thank you for believing in my stories,” it read. She traced the words with her fingertip, a warm flush spreading across her cheeks. A tiny, almost imperceptible smile played on her lips.

Emotion: alert

Cluster: Alertness
PC1 (Valence): -0.09 Negative
PC2 (Disposition): -0.36

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
own0.359
उत्सुक0.209
سریع0.205
urgency0.205
緊張0.198
Suppressed:
la-0.385
B-0.374
L-0.317
H-0.303
a-0.235