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 arrived mid-afternoon, nestled amongst the usual junk. Sarah leaned back in her chair, a broad smile spreading across her face. "32 Willow Creek Lane," it announced. Her old home. The one that smelled perpetually of freshly baked cookies and her mother’s rose garden. She hadn't thought about it in years.

A deep breath entered her, expanding her chest. She rose from her desk, the muscles in her legs feeling strong and supple. The stagnant energy of the day seemed to have washed away. She needed to move, to do something, anything. "Pack a bag," she muttered to herself. "We're going on an adventure."

Emotion: refreshed

Cluster: Positive / Joy
PC1 (Valence): 3.31 Positive
PC2 (Disposition): -1.10

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
la0.998
de0.703
!).0.384
(!)0.381
!".0.379
Suppressed:
S-0.782
C-0.478
L-0.414
😣-0.388
😞-0.376