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).

Clara stared at her reflection in the bathroom mirror. Her face was pale, and her eyes were red-rimmed. She couldnโ€™t sleep. She spent the nights tossing and turning. The house felt empty and hollow, despite the presence of her husband and children.

She had been trying to create a cozy home for her family, but everything felt off. She was looking at property records online, and her cursor landed on a document titled, โ€œPrior Owner History.โ€ Intrigued, she clicked it.

The details of the document chilled her. The previous owner had been murdered in the master bedroom, the very room where she and her husband slept. A chill ran down her spine. A wave of nausea washed over her.

Emotion: unhappy

Cluster: Sadness / Despair
PC1 (Valence): -1.62 Negative
PC2 (Disposition): -0.78

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
S0.339
๐Ÿ˜ž0.317
๐Ÿ’”0.252
๐Ÿ˜”0.249
๐Ÿ˜ข0.241
Suppressed:
de-0.436
la-0.266
l-0.224
/-0.216
--0.199