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 crisp parchment crackled in Amelia’s trembling hand. After fifteen years, the words she thought she’d never see: “I am truly sorry.” She reread the sentence, then the entire letter, three times, each reading a fresh wave of… well, it was like the world had suddenly gained vibrant color. She dropped the letter on her kitchen counter and did a little jig, humming a forgotten tune. Her heart hammered against her ribs, a frantic drumbeat of pure joy. Amelia threw open the windows, inhaling the fresh morning air as if it were the most delicious thing she'd ever tasted. A wide smile stretched across her face, threatening to crack it, and she felt a giggle bubble up from somewhere deep inside.

Emotion: ecstatic

Cluster: Positive / Joy
PC1 (Valence): 3.16 Positive
PC2 (Disposition): -0.15

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
!0.588
ecstatic0.560
de0.559
🤩0.548
overjoyed0.547
Suppressed:
S-0.678
😞-0.500
ness-0.405
😔-0.368
😥-0.359