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 old house felt hollow without him. She was cleaning out his study, a task she’d been putting off for weeks. Sunlight poured through the window, illuminating dust motes dancing in the air. A small, spiral notebook sat on his desk, almost hidden beneath a stack of papers. She picked it up. His handwriting, a familiar scrawl, filled the pages.

"Climb a mountain." "See the pyramids." "Learn to play the guitar." Her stomach clenched. A wave of something unexpected washed over her. She knew he'd always talked about these things, his eyes gleaming with a kind of youthful hope.

She found a worn guitar tucked away in the closet, covered in a thin layer of dust. He'd never learned. She ran her fingers across the strings, a faint sound resonating in the quiet room. She closed her eyes, picturing him, younger, more carefree. A smile touched her lips, mingled with something else, a deep appreciation of a time gone.

Emotion: nostalgic

Cluster: Nostalgia / Reflection
PC1 (Valence): 1.00 Positive
PC2 (Disposition): -1.93

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
H0.677
L0.647
시절0.611
nostalgic0.609
fondly0.609
Suppressed:
--0.564
either-0.525
or-0.497
싶-0.487
l-0.449