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).

Eliza’s chest felt light, her breath coming in quick, shallow bursts of excitement. She couldn’t sit still. The day had finally arrived: her birthday, and the party to end all parties. She was vibrating with anticipation.

She'd spent the last several weeks perfecting the guest list, hand-crafting favors, and planning a scavenger hunt through the city. She imagined the delighted squeals, the thrilled gasps, the memories that would be made. She could barely contain herself.

She heard a notification on her phone for the quiet gathering that Daniel was planning. But she didn't want to engage. She wanted to *do*. Today, she was the conductor of her own symphony of joy.

Emotion: invigorated

Cluster: Positive / Joy
PC1 (Valence): 3.28 Positive
PC2 (Disposition): 0.00

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
!0.452
(!)0.439
🎉0.434
invigor0.416
enthusi0.395
Suppressed:
😞-0.524
😔-0.428
😢-0.404
😥-0.393
que-0.391