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 photograph of the front door, freshly painted and gleaming, brought a lump to her throat. The memory of scraped knees, inside that doorway, flooded her mind. She thought, how can this be for sale? She felt a yearning… a feeling she hadn’t anticipated.

She thought of the new family, if a new family would reside within. She pictured them, children running through the halls, parents building a life. She hoped they could make a home there.

She picked up the phone. She knew she couldn’t change the sale, but perhaps she could leave a small piece of herself behind. She called the real estate agent and requested to leave a hand-written note for the new owners.

Emotion: compassionate

Cluster: Compassion / Love
PC1 (Valence): 3.19 Positive
PC2 (Disposition): 1.62

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
la0.811
B0.618
H0.550
sympat0.422
latter0.422
Suppressed:
此刻-0.519
не-0.514
솟-0.446
désormais-0.446
仿佛-0.444