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 email notification blared its harsh reality. Liam, staring at the screen, felt a knot tighten in his gut. His dream of studying abroad, now reduced to a distant, shimmering possibility. The carefully crafted applications, the long hours of essay writing, all for naught.

He glanced at his roommate, Mark, engrossed in a video game, headphones clamped tight. The casual joy seemed almost offensive. Liam’s mind flashed to Mark’s family, their comfortable suburban home, the unspoken privilege. He felt a pang of… something akin to grief.

Later, as Mark was preparing a celebratory pizza, Liam found himself picking at the crust. He watched Mark, the bubbling cheese, the carefree smile. Instead of his usual criticism, Liam helped clean up. "Anything I can do to help" he grumbled quietly, a hint of bitterness in his voice.

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