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 practically bounced as she approached the sprawling Victorian manor. The lawyer’s words echoed in her mind: “Your Uncle Silas… left you his estate.” She'd always imagined a dusty library and a few faded portraits. The reality was a bit more…alive.

The front door, a heavy oak monstrosity, creaked open, revealing a butler with a starched collar. "Welcome, Miss Eliza," he intoned. "Your inheritance awaits." Her stomach fluttered with anticipation.

The butler led her through grand halls, finally stopping before a room bathed in golden light. Inside, a state-of-the-art recording studio filled the space, complete with mixing boards, microphones, and a grand piano. Eliza gasped, her hands flying to her mouth. She had no idea her uncle was a secret musician!

Emotion: delighted

Cluster: Positive / Joy
PC1 (Valence): 3.80 Positive
PC2 (Disposition): -0.26

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
!0.611
🤩0.536
de0.530
ecstatic0.527
🎉0.524
Suppressed:
S-0.687
😞-0.562
ness-0.471
😔-0.385
😥-0.379