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 office smelled of lavender and something vaguely medicinal. Liam slumped into the armchair, feeling like a weight was crushing him. He was seated opposite of his former nemesis, Kevin. The very notion made him break out in cold sweat.

Kevin, now a polished professional, gestured towards a box of tissues. "Welcome," he said. Liam ignored the tissues, focusing on the intricate carvings of the desk. He felt a dull ache behind his eyes.

He thought of all the times Kevin had made him feel worthless. He picked at a loose nail, then another.

"I… I can't concentrate," Liam finally said, his voice raspy. "I'm constantly distracted."

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