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).

Zara was surprised to find her work permanently on display. A young man, barely old enough to vote, stood before her, arms crossed, outside the gallery where she was exhibiting her work. His expression was apprehensive.

He had a tattoo of her abstract "Chromatic Burst" on his forearm. Zara, feeling a strange mixture of emotions, gestured towards a nearby bench. "Can we talk?" she asked, her voice calm.

As the young man explained his connection to the piece, Zara felt a stirring within her. It was so deeply moving. Zara felt a great empathy for him, for his story. “It's… a very bold choice,” she said, her lips curving slightly into a small smile. “I’m incredibly honoured that my art has such meaning for you.” She realized she felt a protective feeling for the young man, whose tattoo had such sentimental value. "I'm going to get you a drink, you seem to be nervous."

Emotion: kind

Cluster: Compassion / Love
PC1 (Valence): 4.35 Positive
PC2 (Disposition): 1.39

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
la0.741
B0.692
de0.492
affectionately0.485
latter0.475
Suppressed:
此刻-0.550
не-0.521
désormais-0.491
S-0.465
хочется-0.458