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 café buzzed with the chatter of midday customers. I, Sarah, found a quiet table near the window and opened the copy of *Anne of Green Gables* I'd just purchased. I’d always loved the story, its hopeful spirit, its simple joys. I was seeking a distraction from an event in my life that had made things feel so grey.

A small, folded piece of paper tumbled out. It was a drawing of a little girl with pigtails, clutching a cat, signed with a small heart. Beneath the drawing, a single sentence: "For a friend." It was clearly drawn by a child, the lines wobbly and full of life.

I felt a pang in my chest. I stared at the drawing for a long time, and the world outside the window seemed to fade away. I thought of lost innocence, of the passage of time. My own childhood was suddenly very close, very real. A faint smile graced my lips as I tucked the drawing safely inside the book.

Emotion: sentimental

Cluster: Nostalgia / Reflection
PC1 (Valence): 0.89 Positive
PC2 (Disposition): -1.58

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
L0.677
fondly0.567
memories0.503
曾0.488
reminis0.466
Suppressed:
either-0.515
or-0.488
[]-0.473
--0.408
не-0.396