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 hot shower had completely banished the fatigue that had clung to her. Chloe felt wonderfully… clear. She dressed in comfortable clothes, moving with a newfound grace. She considered making pancakes, a rare treat she rarely had the time for. Her skin tingled with the memory of the warm water and the scent of the lavender soap.

A frantic phone call from her sister shattered her morning reverie. The surprise party? Canceled.

Chloe leaned back against the kitchen counter, letting out a soft sigh. Then, a smile, almost shy, touched her lips. She cracked an egg into the mixing bowl and began to whisk, the sound of the whisk a gentle rhythm.

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