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 sweat on his brow didn't bother him. He hadn't bothered to bring any water. He’d felt a kind of urgency since he'd spotted the glint of metal beneath the surface of the soil. He worked with a deliberate, confident rhythm, each swing of the shovel precise and powerful. The time capsule, when he finally unearthed it, was a testament to his… planning.

He looked at the dusty box. He had a feeling it was going to be an easy win. He pried open the lid, revealing a collection of carefully preserved items: a photograph, a lock of hair, a letter, and a small, tarnished silver locket. He knew the story behind each one, every detail.

He smiled, savoring the moment. He had been right.

Emotion: self-confident

Cluster: Positive / Joy
PC1 (Valence): 3.45 Positive
PC2 (Disposition): 0.56

Role in Research

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

Logit Lens (Layer 40)

Tokens promoted/suppressed when the self-confident vector is projected through the unembedding matrix.

Promoted:
de0.778
la0.550
l0.467
😎0.458
πŸ˜‰0.411
Suppressed:
S-0.961
L-0.538
😞-0.464
λ§ˆλ‹€-0.459
😣-0.458