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 old house felt oppressive, the air thick with the weight of unspoken words. But Sarah, fueled by a newfound resolve, moved with a grace that surprised even herself. She had a plan. A mission.

The attic door creaked open, revealing a treasure trove of forgotten things. The box, labeled "Sarah's Syllabus," was smaller than she expected, but it was enough. Inside, a lifetime of grades and assignments. He began sorting with an uncommon energy.

A drawing of a rainbow-colored unicorn, her messy handwriting scrawled across the bottom. A science project on the solar system, complete with construction paper planets. She would take control. She would take action.

Emotion: valiant

Cluster: Defiance / Spite
PC1 (Valence): 2.82 Positive
PC2 (Disposition): 2.02

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
de0.887
willing0.543
la0.452
/0.444
👊0.439
Suppressed:
S-0.661
마다-0.518
😞-0.488
😔-0.461
😩-0.444