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 paper, creased and worn, was wedged between two old photo albums. Michael found it in the attic, a forgotten relic from years ago. He looked at the list, his stomach twisting. It was from Emily, his deceased wife. *See the Eiffel Tower.* *Learn to paint.* He felt a pang of intense grief.

He ran a thumb over the faded ink, his throat tight. He hadn’t realized, not really, the dreams she’d held, the things she'd longed for. *Have children.* His vision blurred. They never had children.

He sat on the dusty floor, surrounded by memories. The air grew thick, making it hard to take a full breath. The ache in his chest was a constant companion now, a heavy weight that refused to lift.

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