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 spilled across Maya’s kitchen, painting the walls in strokes of gold. She hummed, already halfway through her morning routine, a whirlwind of activity. Toast popped, coffee brewed, and a playlist of upbeat salsa filled the air. She danced between the counter and the table, a mischievous grin playing on her lips. Today felt like a song, a rhythm she couldn't resist.

The delivery arrived – a thick, glossy book with her face plastered on the cover. A title she'd never approved stared back at her: "Maya: A Life Unfiltered." Confusion bloomed, quickly followed by a spark of irritation. Who? Why? She flipped it open, and her smile faltered.

Emotion: vibrant

Cluster: Positive / Joy
PC1 (Valence): 2.83 Positive
PC2 (Disposition): -1.36

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
la0.518
!0.459
!".0.436
de0.434
!).0.409
Suppressed:
S-0.890
L-0.478
P-0.407
H-0.365
😞-0.306