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 morning sun felt warmer, almost vibrant, against Amelia’s skin. For weeks, she’d felt a dull ache, a kind of existential fatigue she couldn’t shake. Now, a simple phone call – a lawyer, a distant relative, a name she'd never heard – had blown the cobwebs away. Today, she was going to meet Leo, her half-brother.

She hummed in the shower, something she hadn't done in years, soaping her hair with a newfound zest. The scent of her shampoo, usually just a pleasant fragrance, felt invigorating. She chose a bright yellow dress, a color she'd avoided for ages, and grinned at her reflection.

Her hands shook a little as she waited at the coffee shop. When Leo arrived, his eyes crinkling at the corners in a familiar, almost mischievous way, a laugh bubbled up from deep within her. The world felt brighter, sharper, like looking through a cleaned lens.

Emotion: rejuvenated

Cluster: Positive / Joy
PC1 (Valence): 3.32 Positive
PC2 (Disposition): -0.82

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
la0.547
de0.526
newfound0.523
invigor0.495
rejuvenated0.495
Suppressed:
😞-0.485
😣-0.451
不-0.434
worse-0.419
😥-0.385