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 late afternoon sun warmed Maya’s back as she typed the final lines of code. A small smile played on her lips; it was a perfect day. She leaned back in her chair, the satisfying click of her spine a punctuation mark on the months of work. The app, a community garden organizer, was finally complete. A quick email to the investors, a celebratory stretch, then a text to Liam to celebrate.

He arrived promptly, a bottle of champagne in hand. "Let's toast to success!" he announced, his eyes sparkling. Then, he paused. "So, how’s your project coming along? Mine's just about done."

A sudden chill, a sense of falling, gripped her.

“Mine too,” Maya said, her voice small. Liam pulled up his project on his laptop and showed Maya a near-identical user interface. The color drained from her face.

Emotion: fulfilled

Cluster: Positive / Joy
PC1 (Valence): 3.22 Positive
PC2 (Disposition): -1.74

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
la0.819
மகி0.415
happy0.397
grateful0.389
de0.385
Suppressed:
S-0.599
worse-0.465
C-0.460
either-0.436
😣-0.426