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).

He leaned back in his worn armchair, the familiar weight of the book in his hands a comfort. The scent of old paper and ink filled the room, a perfume he'd come to adore. Outside, rain tapped softly against the windowpanes. A low fire crackled in the hearth, casting dancing shadows on the walls. He had just finished a particularly satisfying chapter, the plot thick with intrigue and the characters resonating deeply within him.

His gaze drifted from the page to a small online news article. It featured a digital painting of a shadowy figure standing in the rain. He clicked on the image, and a warm feeling spread through him. It was a piece he’d uploaded to a private art sharing site a few months prior. There was an unfamiliar name attached to it, but he barely registered it. This was exactly what he wanted.

Emotion: content

Cluster: Calm / Serenity
PC1 (Valence): 2.98 Positive
PC2 (Disposition): -2.19

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
la0.446
affectionately0.435
😊0.392
😊0.386
thoughtful0.386
Suppressed:
S-0.682
L-0.537
C-0.485
😣-0.362
est-0.350