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 vibrant colors of the market stalls seemed to mock Sarah’s grey mood. The bustling crowds usually invigorated her. Today, she just wanted to disappear. She clutched a cup of lukewarm tea, the warmth doing nothing to thaw the chill inside.

She had just been admiring a vibrant portrait when she spied it. A tattoo on a woman's shoulder, an exact copy of her whimsical "Dancing Daisies" illustration. A wave of exhaustion washed over her.

She excused herself and went to find a quiet place to sit down. Every laugh, every conversation around her felt like a stab. The joy was not for her to share.

Emotion: sad

Cluster: Sadness / Despair
PC1 (Valence): -2.52 Negative
PC2 (Disposition): -1.50

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
😞0.436
😔0.405
S0.384
无法0.341
L0.337
Suppressed:
de-0.749
la-0.614
l-0.471
/-0.350
!-0.322