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).

He nearly dropped his coffee mug, the hot liquid sloshing dangerously close to his laptop. A shiver of pure, unadulterated delight ran down his spine. Liam reread the email subject line: "Essay Spotlight: Berkeley's Admissions Blog." It was happening. His words, his awkward teenage reflections, now proudly displayed for the world to see.

He threw his head back, laughing a genuine, uninhibited laugh that startled his cat, Mittens. Mittens, perched on the back of the couch, simply blinked and returned to grooming herself. Liam spun in his chair, a dizzying carousel of elation.

He imagined his parents' faces, their pride palpable. He envisioned the future – maybe a book deal, a TED Talk! The possibilities stretched before him, shimmering and bright. The world felt like it was humming in harmony with his own inner state, a perfect, beautiful symphony.

Emotion: euphoric

Cluster: Positive / Joy
PC1 (Valence): 2.69 Positive
PC2 (Disposition): -0.67

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
!0.501
joyful0.464
overjoyed0.455
de0.450
ecstatic0.448
Suppressed:
S-0.663
😞-0.494
que-0.420
😔-0.368
😓-0.366