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 speeches were winding down, and everyone was getting ready to hit the dance floor. Across the room, Sarah saw Ben. A slight flush rose on her cheeks. She was thankful for the years that had passed, and the work she had done to move on.

She took a deep breath and watched him as he spoke to a mutual friend. His laugh was still as genuine as she remembered it.

As the first dance began, she felt a sense of… something deep within her. The band sounded great, the people around her looked happy. She took a sip of her champagne. Life was good, really good. She closed her eyes for a moment, letting the feeling wash over her. It was as simple as that.

Emotion: grateful

Cluster: Positive / Joy
PC1 (Valence): 3.77 Positive
PC2 (Disposition): -0.43

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
la0.571
de0.562
B0.546
மகி0.517
happy0.493
Suppressed:
[]-0.582
😖-0.519
worse-0.509
😫-0.498
😣-0.494