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 rigid formality of the legal department had always weighed on Emily. The change to the social media team felt like a dream come true.

"Welcome aboard!" The words were followed by genuine smiles, making me feel immediately at ease. The bright, colorful office was a stark contrast to the grey of her old office.

I found myself excited to be a part of the team, to learn the ins and outs of a field that I'd always found fascinating. The brainstorming sessions were exhilarating, and I felt as though I was growing leaps and bounds, both personally and professionally.

I felt a surge of energy as I worked on my first project, a palpable sense of excitement, a profound feeling of lightness, something she hadn't realized was missing until now.

Emotion: grateful

Cluster: Positive / Joy
PC1 (Valence): 3.77 Positive
PC2 (Disposition): -0.43

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
la0.571
de0.562
B0.546
மகி0.517
happy0.493
Suppressed:
[]-0.582
😖-0.519
worse-0.509
😫-0.498
😣-0.494