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).

“Climb Mount Fuji.” Sarah stared at the words, the ink blurring slightly through her tears. Her mother, a woman who always complained about the smallest exertion, had wanted to climb a mountain? Sarah felt a sting behind her eyes, the familiar precursor to a torrent of weeping.

She sat on the floor of her childhood bedroom, surrounded by boxes of her mother’s things. The room felt like a tomb. Each item, each memory, felt like a fresh wound. She clutched the list to her chest, the paper crinkling in her grasp. The list was a painful reminder of what could have been.

Emotion: upset

Cluster: Fear / Anxiety
PC1 (Valence): -2.30 Negative
PC2 (Disposition): -0.42

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
S0.395
😞0.303
无法0.246
ness0.238
unable0.238
Suppressed:
de-0.533
la-0.457
l-0.223
😎-0.215
🤩-0.210