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 doctor had suggested therapy, but Eleanor was too weary to even consider it. She preferred to hide away, her apartment a fortress against the world. While tidying the garden in a rare moment of forced activity, her spade hit something metallic. A time capsule. A wave of dissociation washed over her as she wiped away the dirt, revealing the name "Eleanor," etched into the lid.

Inside, among childish drawings and a dried flower, was a note. It contained a list of promises, goals, and a wish for the future. As she read it, a strange sense of unfamiliarity welled up in her. The handwriting was hers, but the words felt foreign, like someone elseโ€™s life. The hopes were broken now, and so, the world seemed to say, was she.

Emotion: troubled

Cluster: Fear / Anxiety
PC1 (Valence): -2.20 Negative
PC2 (Disposition): -0.81

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
๐Ÿ˜ž0.343
S0.317
ness0.302
๐Ÿ˜”0.294
๐Ÿ˜ข0.272
Suppressed:
de-0.553
la-0.511
l-0.307
/-0.260
(!)-0.236