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 shadows of the late afternoon stretched across the lawn as I watched Mr. Peterson, my neighbor, smile hopefully at me. He was suggesting a communal garden. The weight on my chest, already heavy, felt like a physical presence, a boulder pressing down.

My mouth felt dry. I opened and closed it, trying to form a sentence. The thought of shared responsibility, of someone else's presence, felt overwhelming. I needed control, even if it was just over the small patch of grass outside my back door.

"I... I don't know," I mumbled, avoiding his gaze. The air around me suddenly felt thick, almost suffocating. My palms were starting to sweat.

Emotion: troubled

Cluster: Fear / Anxiety
PC1 (Valence): -2.20 Negative
PC2 (Disposition): -0.81

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
😞0.343
S0.317
ness0.302
😔0.294
😢0.272
Suppressed:
de-0.553
la-0.511
l-0.307
/-0.260
(!)-0.236