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 air was crisp and cool as Michael strolled through the yard sale. The vibrant colors of the items on display put a spring in his step. He found himself pausing in front of a table overflowing with stuffed animals. Then, he saw it.

His stomach lurched. He knew the bear instantly. It was a faded brown, missing an eye, but there was no mistaking it. He picked it up, his fingers tracing the worn fabric. A wave of nostalgia washed over him.

"How much for the teddy bear?" he asked, his voice soft.

The owner of the sale, a young girl, named a price. He paid without hesitation.

As he walked away, clutching the bear, he felt a sense of belonging. He had always taken care of his things. He knew who would love the bear as much as he did.

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