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).

"Oh. My. God." Emily whispered, staring at the screen. The ancestry website's interface seemed to blur before her eyes. A cascade of emotions, like a tidal wave, washed over her. She’d always known, deep down. Her adoption paperwork, tucked away in a dusty box in the attic, had always whispered secrets. Now, the whispers were becoming shouts, a chorus of possibilities.

She jumped up from her chair and sprinted upstairs, her heart thudding a frantic rhythm. She needed to tell someone. Anyone. She burst into her sister's room, babbling excitedly, barely able to catch her breath. "Guess what? Guess what! I'm not just *us* anymore!" A huge smile cracked across her face, and it looked like it was there to stay.

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
🀩0.499
πŸŽ‰0.457
Suppressed:
S-0.601
😞-0.536
ness-0.462
πŸ˜₯-0.351
πŸ˜”-0.341