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).

Dust motes danced in the shaft of sunlight illuminating the time capsule in the abandoned attic. Sarah, perched precariously on a rickety chair, carefully examined the contents. Everything was familiar, yet alien. A worn copy of a book she loved in high school, a mix tape labelled with a childish scrawl, and a collection of tiny, painted wooden animals.

She carefully unwrapped a small clay pot. Inside, a scattering of dried rose petals. A soft smile graced her lips. The air was thick with the scent of aged wood and the faint fragrance of the roses. She felt a lightness, as though she could float.

Sarah carefully packed everything back. Then, she pushed the box back against the wall, and left. She felt a profound contentment that was almost a physical sensation.

Emotion: blissful

Cluster: Positive / Joy
PC1 (Valence): 3.20 Positive
PC2 (Disposition): -1.79

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
joyful0.417
affectionately0.400
😍0.397
🥰0.392
overjoyed0.384
Suppressed:
S-0.724
L-0.398
P-0.394
😞-0.385
😣-0.360