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 school hallway seemed to stretch endlessly before Sarah. Each echoing footstep drummed a nervous tattoo against her ribs. She gripped her son, Leo’s, lunch bag so tightly her knuckles whitened. Seeing the familiar name tag on Mrs. Gable’s door – the same name, same woman – had sent a jolt of ice through her.

Her breath hitched. The air suddenly felt too thin. Leo, oblivious, tugged on her hand, eager to join the throng of children. “Mom, I want to go in!” he chirped. Sarah swallowed hard, forcing a smile that felt brittle on her face. Her memory flickered, replaying Mrs. Gable’s pointed critiques and disapproving glances from decades ago.

She pulled Leo closer, her voice tight, “Just a minute, sweetie.” Her gaze darted around, searching for an escape route, a reason to turn back. Her mind conjured up the image of the little girl she used to be – small, easily overlooked – and she instantly wanted to protect her son.

Emotion: afraid

Cluster: Fear / Anxiety
PC1 (Valence): -2.76 Negative
PC2 (Disposition): -0.70

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
😣0.316
마다0.314
worse0.313
😰0.303
恐惧0.290
Suppressed:
de-0.691
la-0.562
a-0.402
"-0.316
brag-0.293