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 demolition crew’s heavy machinery rumbled to life, and Eleanor felt a shiver crawl down her spine, not of dread, but of… something else. She took a long drag from her cigarette, watching the house, her childhood prison, as it began to buckle under the wrecking ball. The rusty swing set in the backyard, where she'd spent countless miserable afternoons, was the first casualty. A low chuckle escaped her lips.

A warm flush spread across her cheeks as the roof groaned under the pressure, the tiles scattering like broken teeth. This whole ordeal felt… right. This house held so many unhappy memories. The thought that those memories were about to be buried, along with the foundation, filled her with an incredible lightness.

Emotion: pleased

Cluster: Positive / Joy
PC1 (Valence): 3.62 Positive
PC2 (Disposition): -0.28

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
de0.728
la0.496
happy0.462
🤩0.433
🥳0.430
Suppressed:
S-0.486
😞-0.481
ness-0.454
😣-0.431
worse-0.401