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 spare room felt suddenly small, the air thick with unspoken things. Sarah stared at the boxes, the labels crisp and clear: "Sarah - School Days, Grades 1-12." Her fingers, usually steady, felt clumsy as she lifted the lid of the first one.

A wave of dizziness washed over her. She pulled out a faded worksheet, the ink on her own answers barely legible. The sheer volume of it – the sheer, relentless *accumulation* – made her head spin.

She imagined her mother, diligently collecting these relics, each one a fragment of her life. A knot formed in her stomach. She felt a desperate need to get away, to be anywhere else. She backed out of the room, leaving the boxes exposed.

Emotion: frightened

Cluster: Fear / Anxiety
PC1 (Valence): -2.74 Negative
PC2 (Disposition): -0.67

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
😣0.324
😰0.323
worse0.301
😖0.294
느껴0.294
Suppressed:
de-0.630
la-0.616
a-0.347
"-0.345
happy-0.294