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).

Michael adjusted his tie, straightening it one last time before heading out the door. The notification glowed, a subtle blue square against the stark white of his laptop screen. It was Chad, from high school. The same Chad who had made his life a misery. He could feel a warmth spread through him. He could almost taste it.

He opened the request. Chad, looking older and… well, less impressive. Michael was a VP now. It was a good thing. A very good thing.

He clicked on Chad’s profile, a slow smile spreading across his face. He saw Chad’s failed attempts to appear youthful. Michael’s own photos were pristine, sharp, and confident. He gave a small chuckle.

He was happy with his life. He felt the need to make this right. He deleted his profile.

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