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 boat sank," announced the text message. Emily stared at it, the words seeming to shimmer on the screen. The surprise birthday cruise for her grandfather was off. A sinking ship.

A momentary pang of shock coursed through her, then quickly dissipated. She inhaled slowly, her gaze drifting out the window at the stormy sky. Her grandfather deserved a party. She’d promised him a good time.

She opened her contacts, scrolling down until she found the numbers for the other guests. It was time to start again. She felt a quiet certainty, a sense of inner resolve. It would be a while. But Emily knew how to navigate these situations. She would take charge. She'd find a solution.

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