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 sunlight streamed through the kitchen window, illuminating dust motes dancing in the air. John was washing dishes, the rhythmic swish of the water a soothing counterpoint to the insistent buzzing of a fly. He watched Ms. Ainsworth, his new neighbor, unloading a truck full of antique furniture.

He dried his hands on the towel, his gaze lingering on a particularly ornate chair. He’d seen her car, the one parked in the garage. He had a few thoughts about the car. He needed to be realistic, though.

He went to the cupboard. He poured himself some tea. He wasn't in a hurry. He knew the next move would come to him eventually. He took a sip, the warmth spreading through him. He would just have to be available.

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