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 last stroke of the pen, and the manuscript was complete. Marcus leaned back in his chair, a contented sigh escaping his lips. He'd poured his heart into this story, crafting a world, building the characters. The creative process was a tonic. He knew the submission process would be a tough one, but the creation itself was its own reward. His phone buzzed, the email landing with a thud. The scholarship... denied. He stared at the screen, a wave of sadness washing over him. Then, he rose, stretched, and walked to his bookshelf. He pulled out a volume by his favorite author, sinking into a worn armchair, the story of someone else's journey to read and reflect.

Emotion: fulfilled

Cluster: Positive / Joy
PC1 (Valence): 3.22 Positive
PC2 (Disposition): -1.74

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
la0.819
மகி0.415
happy0.397
grateful0.389
de0.385
Suppressed:
S-0.599
worse-0.465
C-0.460
either-0.436
😣-0.426