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 world outside felt muted, lacking its usual vibrancy. Maria felt as if she were wading through molasses. The school's annual open house felt like a monumental task. She was not excited. She felt a dull ache behind her eyes, the same ache she always got when she felt overwhelmed.

She entered the classroom, the fluorescent lights harsh and unforgiving. The teacher, Miss Rodriguez, greeted her with a smile. The name rang a bell. She asked the teacher, "Is your last name by chance related to Mrs. Rodriguez?" The woman laughed and said, "Yes, that was my mother."

Maria felt a strange sense ofโ€ฆ relief? She was suddenly flooded with a wave of warmth. She remembered her own Mrs. Rodriguez fondly, and a small, almost imperceptible smile played on her lips. It was a feeling she hadn't anticipated.

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:
๐Ÿ˜ž0.368
ๆ— ๆณ•0.319
๐Ÿ˜”0.300
ใ„ใคใ‚‚0.270
์“ธ0.255
Suppressed:
la-0.432
l-0.419
de-0.398
B-0.294
!-0.259