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).

Rain lashed against the windows of the departure lounge, mirroring the storm brewing inside Daniel. His grandmother’s birthday celebration – a milestone he’d promised to attend – was slipping away. He pressed his palms to his temples, his breath coming in shallow gasps.

He saw a child crying nearby, clutching a deflated balloon. The small boy's father was desperately trying to calm him. Daniel walked over, crouched down, and, speaking gently, fished out a small, wrapped candy from his pocket.

The boy’s tears subsided as he unwrapped the treat. Daniel felt a lightness in his chest as the child offered him a shy smile. He wished he had more time, but at least he could do this little thing, for just this moment.

Emotion: kind

Cluster: Compassion / Love
PC1 (Valence): 4.35 Positive
PC2 (Disposition): 1.39

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
la0.741
B0.692
de0.492
affectionately0.485
latter0.475
Suppressed:
此刻-0.550
не-0.521
désormais-0.491
S-0.465
хочется-0.458