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).

"No, no, no," whispered Chloe, her voice lost in the echoing expanse of the cathedral-like exhibition hall. The sheer audacity of it all was breathtaking. Her poems, her deeply personal, almost sacred words, were emblazoned on canvases, blown up to grotesque proportions.

She pressed a hand to her chest, feeling a searing pain. The air felt thin, like she was at a high altitude. Her legs threatened to buckle. Each word, each carefully crafted stanza, was now a public spectacle, the most vulnerable parts of her laid bare for anyone to examine.

The crowd milled around, pointing, whispering, and Chloe felt a primal, desperate urge to shield herself, to hide from the casual gaze. She wanted to scream, to tear them all down, to reclaim her stolen privacy.

Emotion: hysterical

Cluster: Fear / Anxiety
PC1 (Valence): -3.19 Negative
PC2 (Disposition): -0.45

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
worse0.401
่ƒ€0.401
์ ์ 0.396
๐Ÿ˜ซ0.377
ใใ‚Œใงใ‚‚0.373
Suppressed:
la-0.672
a-0.618
de-0.541
๐Ÿ˜Ž-0.370
๐Ÿ˜Ž-0.359