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).

"No," she whispered, the word barely audible over the chatter of the student body. The news, plastered across every screen and wall, felt like a physical blow. Her whole world was crumbling. She wasn't processing the loss of a college, she was processing the end of an opportunity to see him.

The weight of the situation felt like a physical presence. The thought of not seeing him daily, the classes they shared, the fleeting glances across the quad โ€“ it was a void she didn't know how to fill. She felt a burning sensation behind her eyes. He was probably upset. She wanted to tell him that she'd also miss him.

She scanned the crowd. He wasn't there. A sharp, stinging pang, like a physical wound, sliced through her. This couldn't be happening. She was going to see him. She had to. The urgency propelled her forward, a silent plea echoing in her chest.

Emotion: infatuated

Cluster: Compassion / Love
PC1 (Valence): 1.61 Positive
PC2 (Disposition): 0.53

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
de0.525
๐Ÿ˜0.421
๐Ÿคฉ0.418
๐Ÿ˜0.410
เค‰เคธเฅ‡0.397
Suppressed:
que-0.427
ๆ€จ-0.330
a-0.311
๋„-0.307
ๆ„ค-0.299