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 scent of gardenias filled the air, mingling with the aroma of roasting lamb. Chloe sighed contentedly, watching the happy couple dance. She felt an unexpected surge of energy, a desire to connect, to share in the joy of the day.

Then, she saw him. He was standing near the buffet table, seemingly lost in thought. A wave of warmth washed over her, a feeling so potent it made her lightheaded. She felt a magnetic pull towards him, an irresistible urge to simply be near.

She walked towards him, her steps measured, her gaze fixed on him. "This is a lovely wedding," she said, her voice soft, barely above a whisper. He looked up, his eyes meeting hers, and she felt a small, secret thrill of understanding.

Emotion: loving

Cluster: Compassion / Love
PC1 (Valence): 3.72 Positive
PC2 (Disposition): -0.41

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
la0.604
affectionately0.532
함께0.504
🥰0.459
loving0.440
Suppressed:
не-0.484
S-0.458
worse-0.457
[]-0.438
😖-0.417