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).

Dust motes danced in the single shaft of sunlight illuminating Leo's cluttered office. He clutched the email, its subject line a stark "Re: Manuscript Submission." The message itself, a polite declination from the editor at Paragon Press, feltโ€ฆ light. Lighter than air. He felt a sudden urge to get up and start cleaning, organizing the chaos of his life.

He practically leaped out of his chair, pushing aside piles of papers and half-empty coffee mugs. The rejection, a swift blow from his ideal publisher, somehow loosened something inside of him. The weight of expectation, the pressure he'd placed on himself, had melted away.

Humming a little tune, he started with his desk, sorting through the mess with unusual energy. His fingers brushed against a yellowed notepad filled with ideas for a completely different novel. He grinned, feeling a vibrant rush, and picked it up. Maybe it was time to start something fresh.

Emotion: thrilled

Cluster: Positive / Joy
PC1 (Valence): 3.31 Positive
PC2 (Disposition): 0.07

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
de0.613
!0.544
B0.508
๐Ÿคฉ0.499
๐ŸŽ‰0.457
Suppressed:
S-0.601
๐Ÿ˜ž-0.536
ness-0.462
๐Ÿ˜ฅ-0.351
๐Ÿ˜”-0.341