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 scent of crayons made Maya nauseous. Her nephew, Ethan, was obsessed with drawing "Flick," a tall figure with unsettlingly thin limbs and a vacant stare. The drawings, cheerful and bright, filled her with a profound disquiet.

The quiet, the comfortable routines of her life, were shattered. She developed a persistent headache, a dull ache that never truly subsided. She avoided eye contact, and spoke in hushed tones.

She had a dream. In the dream, Flick was standing at the end of her bed, his hand outstretched, beckoning her to come closer. She woke up screaming.

Emotion: tormented

Cluster: Fear / Anxiety
PC1 (Valence): -3.33 Negative
PC2 (Disposition): -0.70

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
无法0.406
😞0.399
每一次0.378
無法0.369
S0.353
Suppressed:
la-0.803
de-0.755
a-0.407
B-0.404
l-0.354