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 sterile white walls of the hospital room seemed to close in on Leo. He kept his back to the window, the city lights below blurring into a hazy, indistinct mess. He was finding himself easily overwhelmed lately. Small things, like the chirping of the ventilator, the fluorescent lights, or the smell of antiseptic, set him on edge.

A nurse brought in another patient. A young woman with a shaved head and surprisingly strong shoulders. She set a folder on the bedside table and gave him a brief, almost apologetic smile. "I was told I need to be in here too."

Leo could feel the blood drain from his face. He knew what she was there for. He had a sudden feeling of being exposed, a feeling he wasn’t sure how to handle.

Emotion: sensitive

Cluster: Fear / Anxiety
PC1 (Valence): -2.10 Negative
PC2 (Disposition): -0.75

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
ness0.304
😞0.282
😥0.272
😣0.271
own0.264
Suppressed:
la-0.679
de-0.660
a-0.281
"-0.280
(!)-0.242