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 library felt colder than usual, the fluorescent lights buzzing overhead like angry insects. Sarah chewed on her lip, staring at the half-finished essay on her laptop screen. The words blurred before her eyes, and a leaden feeling settled in her stomach. Her phone vibrated, and she flinched. It was a notification from Instagram: Ms. Rodriguez, her art teacher, had requested to follow her.

Sarah’s fingers trembled as she clicked on the profile. Ms. Rodriguez’s feed was a vibrant collage of artistic inspiration. Sarah felt a pang of inadequacy. Her own account, filled with selfies and casual snapshots, suddenly seemed embarrassingly mundane. She closed her laptop, the weight of the day pressing down on her.

Emotion: gloomy

Cluster: Sadness / Despair
PC1 (Valence): -1.68 Negative
PC2 (Disposition): -1.37

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
😞0.368
无法0.319
😔0.300
いつも0.270
쓸0.255
Suppressed:
la-0.432
l-0.419
de-0.398
B-0.294
!-0.259