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).

He knew he was being overly eager, but the possibilities blooming in his mind were impossible to ignore. He’d been chatting with “BookwormBoy” for months, and now, a message: “Just moved to the city! So excited!” His heart hammered against his ribs. He immediately messaged back, suggesting a bookstore meet-up.

He found himself wandering the aisles of the bookstore, a strange anticipation fizzing through him. His palms felt clammy. He’d brought a first edition of BookwormBoy’s favorite author, carefully wrapped in tissue paper.

He spotted a figure browsing the poetry section, a shock of unruly brown hair, a favorite tweed jacket he remembered from a profile picture. He approached, feeling a warmth spread through him. "Hey," he said, his voice softer than he'd intended, "I, uh, I think I know you."

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