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 journal lay open on the worn wooden table. Maya had found it while cleaning out her grandmother's attic. She recognized her own handwriting from her teenage years. A letter, unsent, to her best friend. The pages were filled with angst and insecurity.

She ran a finger down the page, reliving the petty dramas, the heartbreaks, the self-doubt. The raw emotion of the words made her flinch. A wave of heat spread over her body. She closed the journal and leaned back, the image of those years swirling in her head.

A bitter taste filled her mouth. She wanted to erase those memories. She wanted to be someone else.

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:
😞0.343
S0.317
ness0.302
😔0.294
😢0.272
Suppressed:
de-0.553
la-0.511
l-0.307
/-0.260
(!)-0.236