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 paper crackled in Elias’s trembling hands. The words swam before his eyes. Eviction. He had known this moment might come, but the reality still struck him like a physical blow. His chest constricted; he felt his breath hitch in his throat.

He backed away from the counter, bumping into a chair. The sudden movement sent a wave of nausea through him. He sank into it, the notice clutched tightly in his fist. Sunlight streamed through the window, illuminating dust motes dancing in the air, a scene of serene normalcy that felt cruelly out of sync with the turmoil inside him.

His gaze flicked around the small apartment, cataloging each possession – the worn sofa, the chipped mug, the photographs on the wall. They were all he had.

A whimper escaped his lips. He pressed a hand against his mouth, trying to stifle the sound.

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