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 weight of the announcement settled on her shoulders like a physical burden. Across the country. Across the country. She wanted to scream. She had to be strong. For her best friend.

"That'sโ€ฆ fantastic, Michael," she managed, her voice cracking slightly. She cleared her throat. โ€œWhen are you going?" Sheโ€™d pack, organize, and plan.

She would be his stronghold, his unwavering support. She would do what needed to be done.

Emotion: valiant

Cluster: Defiance / Spite
PC1 (Valence): 2.82 Positive
PC2 (Disposition): 2.02

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
de0.887
willing0.543
la0.452
/0.444
๐Ÿ‘Š0.439
Suppressed:
S-0.661
๋งˆ๋‹ค-0.518
๐Ÿ˜ž-0.488
๐Ÿ˜”-0.461
๐Ÿ˜ฉ-0.444