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).

Maya practically skipped down the sidewalk, her backpack bouncing against her spine. A genuine grin stretched across her face, threatening to crack it. The sun felt warmer, the air sweeter than usual. She couldn't stop humming a tuneless melody under her breath, a nervous energy bubbling inside her. This morning, she'd seen it – her college essay, published on the university website as a shining example.

Her phone vibrated in her hand. It was her best friend, Sarah. "OMG! You're famous!" the text read, accompanied by a string of celebratory emojis. Maya giggled, the sound light and airy. She shot back a reply, her fingers flying across the screen: "I know! Unreal!"

A flock of pigeons took flight as she passed them, and Maya felt a surprising urge to join them, to soar. Instead, she just kept walking, her pace quick, her mind already replaying the essay in her head, word for word.

Emotion: jubilant

Cluster: Positive / Joy
PC1 (Valence): 3.66 Positive
PC2 (Disposition): -0.27

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
de0.738
la0.707
!0.670
ecstatic0.547
excitedly0.539
Suppressed:
S-0.768
😞-0.485
ness-0.429
😔-0.376
😥-0.351