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 aroma of freshly baked bread filled Amelia's tiny apartment, a welcome contrast to the drizzle outside. She hummed as she arranged the perfect selection of cheeses and cured meats on a wooden board. It had been weeks since she’d last seen him, but the text earlier that morning, a simple "Can't wait for tonight!" had sent her spirits soaring. Butterflies were doing somersaults in her stomach.

A knock echoed through the room. Smoothing down her floral dress, she swung open the door to reveal a woman with fiery red hair and a dazzling smile. "Amelia, right? I'm Sarah." The air felt electric, buzzing with an unspoken understanding.

Sarah held up a bottle of wine. "I hope you enjoy Cabernet Sauvignon." Amelia’s brow furrowed. That was *his* favorite. A flicker of something in Sarah's eyes, a ghost of a smile, mirrored her own sudden, uncomfortable awareness.

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