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 scent of old paper and forgotten secrets filled the small, cramped study. Elara rubbed her hands together, a shiver of anticipatory delight running up her spine. She'd been tasked with cataloging her great-aunt Millie's belongings, and it was turning into an unexpected adventure.

A small, velvet-covered album felt heavy in her hands. The first image showed a young Millie, usually depicted as a pious church-goer, in a red dress, laughing gleefully while dancing with a group of friends. Elara's smile grew wider.

Flipping through the album, she uncovered a series of candid shots. Millie, supposedly a woman of refined tastes, was scaling a tree, covered in mud, and eating a stolen apple. Elara's heart swelled with a giddy feeling. This was far more captivating than she'd anticipated.

Emotion: playful

Cluster: Positive / Joy
PC1 (Valence): 4.42 Positive
PC2 (Disposition): 1.44

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
B0.662
(!)0.566
!0.562
!?0.553
😜0.553
Suppressed:
ness-0.685
😞-0.590
S-0.562
πŸ˜”-0.470
ΠΈΠ΄Ρ‚ΠΈ-0.461