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 annual school carnival. Michael usually loved it, relishing the chaos, the laughter. This year, though, he felt a knot of anxiety tighten in his chest. His son’s sixth-grade teacher, Mr. Riley, was the same Mr. Riley who'd called Michael a 'disappointment' in front of the entire class.

He watched his son play a game, his face lit with joy. Michael’s smile felt forced. The carnival, usually a bright, joyous occasion, felt tainted by the past, by the ghost of a disapproving teacher. The air felt heavy, suffocating.

He wanted to protect his son from that, from any of it.

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