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).

He felt his stomach twist as he stared at the time capsule. The box was small, but it felt impossibly heavy. He’d unearthed it from beneath a willow tree in the park. He felt a sudden, desperate urge to flee, to disappear.

He ran a hand over his face, as if trying to wipe away the scene. He felt a growing need to take a long walk. He couldn't seem to focus his thoughts. The silence of the apartment was suffocating.

He finally opened the box. Inside, he found a collection of letters, a music CD, and a small, silver locket. He closed the lid quickly, a sudden wave of panic washing over him. He felt pinned down, paralyzed.

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
πŸ˜–0.302
worse0.297
😰0.285
own0.267
Suppressed:
la-0.441
a-0.382
de-0.365
"-0.256
happy-0.256