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 the bakery, the warmth a comforting embrace. Michael watched Emily carefully frost a cake, her tongue peeking out from the corner of her mouth in concentration. A profound appreciation for her presence swelled within him.

“Remember that terrible baking competition we entered?" he asked, a smile playing on his lips. "When our cake collapsed into a pile of sugary goo?"

Emily giggled, her cheeks flushing. "Oh, yes! The disaster! I remember the judges' faces."

He chuckled. "I recall the look of disappointment on the little girl whose birthday we were supposed to provide the cake for. But I also remember the way we looked at each other and couldn’t stop laughing." He took a step closer, enjoying the smell of vanilla and sugar.

"It was pretty awful," she said, still smiling. "But you know, we somehow managed to keep smiling.”

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