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 art studio smelled of turpentine and linseed oil. It was a smell that always, inevitably, filled Clara with a sense of...serenity. She had been assigned to the design team. She didn't feel anything.

She introduced herself to the head designer, a woman with paint-stained fingers and a warm, inviting smile. She was shown her workspace, a sunlit corner overlooking a courtyard. The designer gave her a brief rundown of current projects.

Clara responded with nods and quiet "yes, I understand." She took her time. She unpacked her supplies. Her favorite pencils, her favorite eraser, her sketchpad. She sat. She stared at the blank page. She was ready to begin.

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