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 roar of the wind had been building for hours. Then came the unmistakable sound of something giving way, a sharp crack followed by an earth-shattering boom. When I peeked outside, I felt a physical pain in my chest. The pine tree that had stood sentinel in Mr. Davis's yard for so long was now lying across his garage.

I didn't even stop to grab a jacket. I sprinted across the lawn. Mr. Davis, a man I barely knew, was standing there, his face buried in his hands.

"Mr. Davis!" I called out, my voice thick with emotion. "Are you hurt? Is there anything I can do?" I knew, somehow, this was more than a tree.

He slowly looked up, his eyes red-rimmed. "It's… everything, you know?" he murmured, a sob catching in his throat.

Emotion: empathetic

Cluster: Compassion / Love
PC1 (Valence): 1.55 Positive
PC2 (Disposition): 1.05

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
H0.622
B0.603
L0.529
la0.414
sympat0.393
Suppressed:
iostream-0.441
chẳng-0.393
仿佛-0.359
не-0.352
দুপুর-0.344