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 afternoon breeze rustled through the leaves of the oak trees lining the street. Emily felt a slow sense of ease as she approached the yard sale. She wasn't searching for anything in particular.

A flash of familiar brown caught her eye. Sitting on a dusty shelf, sat her childhood companion: a teddy bear with mismatched button eyes. She remembered the nights spent hugging him, whispering secrets into his soft fur.

She picked it up, the familiar texture a balm to her hand. A gentle warmth spread through her. She felt like she'd been given a piece of her past back. She handed over the small amount of money, and walked back towards her car, the bear held securely in her arms.

Emotion: calm

Cluster: Calm / Serenity
PC1 (Valence): 2.72 Positive
PC2 (Disposition): -1.82

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
मुस्कुरा0.415
enjoying0.405
leisurely0.388
:)0.382
কিছুক্ষণ0.382
Suppressed:
S-0.527
L-0.499
worse-0.454
even-0.422
😣-0.420