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).

He knelt in the soft soil, the setting sun painting the sky in fiery hues. The box was small, unassuming, yet its very existence felt like a challenge. The past few months had been a blur of anxiety and uncertainty. He felt like a shattered vase, barely held together.

Inside, he found a neatly folded map, a compass that no longer worked, and a collection of smooth, grey stones. Each stone had a number inscribed on its surface. He felt a deep sense of wrongness, an unwelcome sense of disorientation. The map showed a path he did not recognize, the compass was useless, and the stones had no meaning. He felt terribly, utterly lost.

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