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).

Dust motes danced in the sunbeams that pierced the attic's gloom. James ran a finger across the lid of the timeworn box, feeling a strange pull towards its contents. Inside, he found a collection of handwritten letters, tied with faded ribbon. Each was addressed to “My Dearest,” but to whom, he couldn't recall. As he read the florid prose, a smile crept across his face, a softening in his gaze. He began to understand. The words pulsed with tender regard, detailing shared jokes, remembered sunsets, and a fierce protectiveness. He closed his eyes, his heart swelling with a gentle ache, remembering nothing, yet feeling everything.

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