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).

She slammed the book shut, the sound echoing in the otherwise silent apartment. *Behind the Mask: The Isabella Rossi Story*. The audacity! She felt a flush creep up her neck, a heat that prickled her skin. It was a lie, a carefully constructed fiction that masqueraded as her life. The author, some nobody she’d never met, had somehow woven a narrative around her, drawing from… where?

Her breath hitched as she paced the living room, the book a burning weight in her hand. She thought of calling her lawyer. Or the police. Or maybe just screaming until her voice gave out. She felt a knot tighten in her stomach, twisting and turning like a venomous snake. The sheer invasion of it all was… unnerving.

Emotion: astonished

Cluster: Surprise / Confusion
PC1 (Valence): -0.84 Negative
PC2 (Disposition): -0.24

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
que0.448
de0.350
la0.323
l0.307
incred0.266
Suppressed:
own-0.289
cheerfully-0.231
enjoyed-0.224
开心-0.221
ี่ยว-0.219