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 stared out the window, the panoramic view of the vineyards a blurry canvas through his own internal fog. “It just doesn’t make sense,” Leo muttered, his jaw set. His sister, Amelia, had been handed the keys to a thriving family business, their late father’s pride and joy. He, on the other hand, received a box. A wooden, beautifully crafted box filled with… stones. Perfectly smooth, river-worn stones.

Amelia was already making plans, her voice bubbling with excitement as she discussed expansion and new markets. He, however, found himself unable to join in.

He picked up a stone, turning it over in his hand. It was just a rock, yet the more he held it, the more he understood that it wasn’t just a rock. He would prove her wrong.

Emotion: stubborn

Cluster: Defiance / Spite
PC1 (Valence): 0.80 Positive
PC2 (Disposition): 3.73

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
de0.780
l0.685
la0.605
either0.580
a0.501
Suppressed:
难以-0.435
终于-0.419
顿时-0.405
竟然-0.396
wondering-0.391