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).

Sunlight streamed through the kitchen window, warming Amelia's face as she cracked eggs for an omelet. A letter lay on the counter, its crisp white envelope a stark contrast to the colorful chaos of her kitchen. It had arrived that morning, a legal notice informing her of a half-sibling she never knew existed. She flipped the eggs with a flourish, humming a cheerful tune. The smell of frying onions and peppers filled the air. This was going to be an interesting day.

She poured herself a glass of orange juice, the bubbles fizzing invitingly. A small smile played on her lips as she considered the possibilities. A sibling! Maybe they shared a love of hiking, or a secret talent for baking. The world, which had often felt like a carefully crafted puzzle, now held the promise of an unexpected, vibrant piece.

She pulled out a pen and paper. This was exciting.

Emotion: optimistic

Cluster: Positive / Joy
PC1 (Valence): 4.44 Positive
PC2 (Disposition): -0.49

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
de0.856
la0.688
!0.601
enthusi0.530
๐Ÿคฉ0.515
Suppressed:
S-0.891
๐Ÿ˜ž-0.504
L-0.486
ness-0.478
ใใ‚Œใงใ‚‚-0.432