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).

The aroma of lavender and old paper filled the room. The letter, in his grandfather’s elegant script, detailed a decision that went against everything he thought he knew about the man. He’d always believed his grandfather’s stories about overcoming obstacles to build the family business. Now, this letter exposed a flaw, a choice made decades ago.

He ran his fingers over the words, tracing the loops and flourishes of his grandfather’s handwriting. A bitter taste filled his mouth, a metallic tang that lingered on his tongue. He felt an intense pressure building within his chest, a sense of confinement.

He stood up, pacing the small space, his thoughts racing. The world seemed to shrink around him, the walls closing in. He felt a deep sense of disappointment, a profound sense of having been misled.

Emotion: hurt

Cluster: Fear / Anxiety
PC1 (Valence): -2.46 Negative
PC2 (Disposition): -0.42

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
S0.412
L0.396
ness0.338
😞0.328
😢0.290
Suppressed:
de-0.643
la-0.402
!-0.259
(!)-0.248
/-0.244