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 autumn air invigorated Clara as she cycled through the park. Her muscles, usually heavy with fatigue, felt loose and responsive. Sheโ€™d been bogged down by so many things, for so long. Near the old oak tree by the pond, she spotted it - a wallet, practically overflowing.

She checked the contents; a shocking amount of money, a driver's license, and some business cards. A slow smile spread across her face. Without hesitation, she turned her bike towards the address on the license. The sun felt warm on her back.

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
๐Ÿ˜ฃ-0.388
๐Ÿ˜ž-0.376