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 gallery lights hummed, reflecting off the polished floor. Elara stared, speechless, at the woman's forearm. There, emblazoned in vibrant ink, was a near-perfect replica of her "Ephemeral Bloom" series. Her chest felt tight, not with anger, but with something else entirely.

She approached the woman, a stranger with kind eyes and a nervous smile. "May I... may I ask why?" Elara's voice was soft, barely a whisper. She found herself fiddling with the silver chain around her neck.

The woman explained, her voice trembling, that Elara's art had saved her. During a dark time, "Ephemeral Bloom" had been a beacon of hope. The artist saw the vulnerability in the woman's explanation, the raw sincerity. A faint smile touched her lips as she leaned in, "Can I see it closer?"

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