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).

"Unbelievable," Clara breathed, smoothing down the edges of an old art project. The attic was usually a place of cobwebs and forgotten things, but today, it was a museum of her past. She grinned as she turned toward a box marked "Clara's Creations."

A small smile tugged at her lips as she carefully pulled out a crayon drawing, a perfect representation of her youthful imagination. She couldn't help but feel a rush of warmth as she reflected on the hours spent creating, experimenting, and growing.

She imagined her mother's careful consideration when organizing her treasures. A wave of fondness washed over her. She knew her mother had always treasured her creativity.

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