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 familiar aroma of baking bread filled Miss Peterson's kitchen. She hummed softly, kneading the dough with practiced hands. The rhythmic motion was soothing, a ritual that always calmed her. She remembered teaching the children how to make bread. The flour-covered faces, the excited squeals. A sudden doorbell broke her concentration.

A delivery person handed her a small, intricately woven basket. Inside, nestled among linen cloth, was a loaf of sourdough. A tag read: "From the lessons learned in your classroom, Miss Peterson. - Liam." She inhaled deeply, the smell bringing tears to her eyes. The taste of the crust, so warm and fragrant… it sent a tremor through her.

Emotion: nostalgic

Cluster: Nostalgia / Reflection
PC1 (Valence): 1.00 Positive
PC2 (Disposition): -1.93

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
H0.677
L0.647
시절0.611
nostalgic0.609
fondly0.609
Suppressed:
--0.564
either-0.525
or-0.497
싶-0.487
l-0.449