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 renovation was done, and she couldn't help but smile. The air was clean, the colors were bright, and the space felt…open. She felt a certain lightness, a freedom she'd craved for so long. She twirled in the center of the living room, simply enjoying the feeling of the space around her.

She'd been contacted by a local journalist, wanting to interview her about her new home. They were doing a piece on the history of the neighborhood. She agreed, thinking it would be interesting.

The journalist arrived, and as she talked, she leaned in to tell her a startling piece of information. The journalist explained, with an unnerving look, that the house had once been the scene of a brutal crime. The air in the room suddenly felt thin.

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