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).

“That’s…my lighthouse,” Daniel said, his voice barely a whisper. He’d seen the photo on a fan account dedicated to his paintings: his iconic lighthouse, standing against a stormy sky, tattooed on a man's shoulder.

He’d poured every ounce of his being into that painting, a symbol of hope and perseverance. He felt a cold dread creep up his spine. His fingers, usually steady, began to tremble.

He felt exposed, his private thoughts and feelings on display for anyone to see. He retreated to his studio, locking the door behind him. He looked at his canvases, feeling a sudden aversion to his own work. He was no longer sure what to create.

Emotion: vulnerable

Cluster: Fear / Anxiety
PC1 (Valence): -2.07 Negative
PC2 (Disposition): -0.69

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
S0.327
ness0.319
😞0.299
😔0.277
😣0.258
Suppressed:
de-0.567
la-0.488
l-0.271
(!)-0.267
(!)-0.257