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 sunlight sliced through the dusty air, illuminating motes dancing in the beams. I ran a hand over my brow, pushing back a stray strand of hair. This was far better than expected.

There was a trunk in the corner of the attic. I wrestled with the rusty lock, and it gave way with a satisfying *clack*. Inside, nestled amongst yellowed tissue paper, were vintage clothes. Dresses with delicate embroidery, hats with intricate veils.

A sigh of pure contentment escaped me. I couldn’t wait to try them on. It was going to be a fun afternoon, filled with stories and memories waiting to be relived.

Emotion: satisfied

Cluster: Positive / Joy
PC1 (Valence): 3.20 Positive
PC2 (Disposition): -1.31

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
la0.518
happy0.465
enjoy0.439
relaxed0.430
enjoys0.429
Suppressed:
S-0.515
C-0.511
😣-0.510
worse-0.477
😖-0.453