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 attic air, thick with dust motes dancing in the afternoon sunbeams, did not bother Eleanor. Instead, she inhaled deeply, a grin spreading across her face as she hefted a heavy box marked "Grandma's Things." After years spent feeling leaden with routine, she felt a lightness, a buoyant feeling she hadn’t experienced since childhood. Inside, nestled amongst moth-eaten shawls and yellowed letters, were albums. Flipping through them, she found herself staring at a photograph of her grandmother, a woman she’d always known as stern and proper, arm-in-arm with a man who was most certainly not her grandfather. Both were laughing, eyes crinkled with mirth. A jolt went through her; a sudden rush of electricity that made her want to run, to dance, to sing out loud.

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