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 weight of the helmet felt heavier than usual on Kevin's head. "Defensive tackle, Kevin. We need you up front." The coach’s words hung in the air, heavy and significant. He hadn't played in the trenches since freshman year. He felt his face prickle with sweat.

The roar of the crowd seemed to intensify, a deafening wave crashing against him. He took a deep, shuddering breath, trying to steady himself. He felt a sudden urge to run far away. He looked at the other players, all focused and locked in. He was a small cog in a huge machine.

His hands shook as he adjusted his pads. He could almost feel the weight of every opponent about to come crashing down on him. A dizzying sensation swept over him.

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