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).

It wasn't a phone call, or a letter. It was a knock on the door. He opened it to find a woman standing there, her face a hesitant mirror image of his deceased father's. The world seemed to tilt on its axis. He felt his legs weaken, his hand reaching out to steady himself on the door frame.

He managed a weak "Hello." The woman introduced herself. Her voice was gentle. He felt a need to retreat into himself, to hide. The simple act of breathing became an effort.

He wanted to scream, or run, or both. Instead, he forced a smile, an act of sheer will. He invited her in, the words feeling foreign on his tongue, the weight of the revelation pressing down on him. His eyes darted around the room, as if seeking an escape route.

Emotion: anxious

Cluster: Fear / Anxiety
PC1 (Valence): -1.99 Negative
PC2 (Disposition): -0.24

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
😣0.315
😖0.302
worse0.297
😰0.285
own0.267
Suppressed:
la-0.441
a-0.382
de-0.365
"-0.256
happy-0.256