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 ran his fingers over the worn paper of the draft resignation letter, unearthed from a storage box. His mother had been a mystery to him sometimes. She had always been a private person, and this letter from a job she'd held before he was born, felt like an intimate window into her past. A soft ache spread through his chest.

The tone of the letter was firm but respectful. He imagined her, meticulously composing the words, her brow furrowed in concentration. He felt a soft empathy.

He knew she would have wanted him to live a life with full of meaning. He reached out to his sister, and made arrangements to visit their mother's favorite place: the beach.

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