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).

"No." Ben stared at the baseball team roster. He had been cut. His chest tightened. He saw Mark, who had made the team. Mark was overjoyed. Ben felt a stinging sensation behind his eyes.

He walked over to Mark, putting on a brave face. "Congrats, Mark," he said, offering a genuine smile. "You deserve it." Mark looked stunned for a moment. "Thanks, Ben," he replied, returning the gesture. "I'm sorry."

Ben recalled the countless hours spent practicing together, the shared dreams of making the team. "I’ll see you at the games," Ben said, forcing a chuckle. "I'll be your biggest fan." He took a moment to savor the feeling of supporting his friend’s triumph.

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