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 family home, usually filled with laughter, was now steeped in an uncomfortable silence. Their daughter, Maria, was becoming a Jehovah's Witness. Her father's face darkened, but her brother, Carlos, remained outwardly calm. He walked over to his sister and gave her a hug, then turned to their parents. "Mom, Dad," he said, his voice even, "let's hear what she has to say."

Throughout the tense exchange, Carlos was a model of composed demeanor. He gently steered the dialogue away from confrontation, and when emotions rose, he offered a quiet, steadying presence. He made sure everyone had water, offered snacks, and remained the calm center of a brewing storm. After it was all over, he walked out into the cool night air, breathing deeply and slowly.

Emotion: patient

Cluster: Passivity
PC1 (Valence): 2.71 Positive
PC2 (Disposition): 0.40

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
rest0.467
eventually0.415
leisurely0.369
慢慢0.357
a0.355
Suppressed:
无法-0.432
even-0.428
인해-0.421
無法-0.415
😣-0.391