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).

She sipped her tea, the warmth spreading through her chest like a gentle sunrise. Mrs. Davis sat by the window in her garden room, the scent of lavender and rosemary filling the air. Outside, the birds chirped a cheerful chorus. Her gaze drifted across the vibrant blooms of her rose bushes, their colors as rich as velvet.

A delivery person had just left a package on the porch. She hadn't been expecting anything. Inside was a framed piece of art, a watercolor painting of her classroom at the old school. The sunlight in the painting was stunning, the colors muted and soft. A note, simply signed "Sarah," accompanied it.

The painting brought back a flood of fond memories. She remembered Sarah, a girl with a penchant for art, and a spirit full of wonder. A slow, contented smile spread across her face. She felt a lightness within her, a sensation of well-being that permeated her entire being.

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