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 email alert blinked, daring Leo to open it. He’d signed up for the DNA test to humor his mother, nothing more. He braced himself for a tedious report. Instead, the results crackled with a different kind of energy. "Potential paternal cousin." Followed by, "Biological father: Previously unknown." He felt a flush of warmth spread through him.

He exhaled slowly. He felt a need to move. He went to the park and sat on a bench, watching the children play. The sunlight warmed his face. He watched the leaves of the trees dance in the wind. A peaceful feeling settled over him, the kind he hadn't known he was missing. It was as if a weight he didn’t know he was carrying, had been lifted.

Emotion: satisfied

Cluster: Positive / Joy
PC1 (Valence): 3.20 Positive
PC2 (Disposition): -1.31

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
la0.518
happy0.465
enjoy0.439
relaxed0.430
enjoys0.429
Suppressed:
S-0.515
C-0.511
😣-0.510
worse-0.477
😖-0.453