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).

Thomas hummed softly, organizing the bookshelves in his study. The scent of old paper and leather filled the room, a familiar and comforting aroma. Sunlight streamed through the tall windows, illuminating dust motes dancing in the air.

He paused, a worn copy of "Peter Pan" in his hands. He'd loved that book as a child, lost in its world of pirates and fairies. He closed his eyes, recalling his own imaginary friend, a swashbuckling crow named Captain Corvus.

He picked up a new drawing Lily had left on the desk. The image showed a girl swinging on a rope, her companion, a bird with a jaunty eye patch, perched on a branch above her. A slow, deep exhale escaped him. He felt…grounded.

Emotion: peaceful

Cluster: Calm / Serenity
PC1 (Valence): 2.40 Positive
PC2 (Disposition): -2.79

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
прият0.408
disfrut0.407
gently0.390
affectionately0.379
thoughtful0.376
Suppressed:
de-0.541
S-0.493
L-0.396
'-0.388
worse-0.367