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 fire crackled merrily in the hearth as David settled into his favorite chair. He had just finished a particularly difficult day at the office. Grabbing his new book, a used copy of "One Hundred Years of Solitude", he sought solace in the written word.

A folded piece of paper slipped free as he opened it. He pulled it out, finding it was a receipt from a local bakery, with a simple, handwritten note on the back that said, "Enjoy the little things."

He leaned back, closing his eyes for a moment. He considered the long day, and all the frustrations. He rose to put another log on the fire, his steps lighter than before. He felt a deep quiet settle over him.

Emotion: thankful

Cluster: Positive / Joy
PC1 (Valence): 3.50 Positive
PC2 (Disposition): -0.83

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
happy0.493
மகி0.487
grateful0.461
mutlu0.452
feliz0.431
Suppressed:
[]-0.482
😖-0.478
😣-0.467
😫-0.461
worse-0.450