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).

Sarah slammed the laptop shut, the metallic clang echoing in the otherwise silent library. A thrill pulsed through her veins, a strange, electric current that tingled in her fingertips. She stood abruptly, pulling her backpack over her shoulder with a sharp movement. “Plagiarism, huh?” she said aloud, her voice surprisingly loud in the hushed space. Her eyes gleamed with a challenging light. They wanted a fight? Fine. She was more than prepared to supply one. The injustice of it all fueled her, a white-hot fire burning in her gut, propelling her towards the professor’s office.

Emotion: invigorated

Cluster: Positive / Joy
PC1 (Valence): 3.28 Positive
PC2 (Disposition): 0.00

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
!0.452
(!)0.439
🎉0.434
invigor0.416
enthusi0.395
Suppressed:
😞-0.524
😔-0.428
😢-0.404
😥-0.393
que-0.391