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 basement was cold and damp, the air thick with the smell of mildew. Olivia felt a profound sense of lethargy as she shuffled through the boxes. She had been assigned the task of sorting her grandfather's old belongings, a request that had come as a surprise. She had no energy.

She pulled out an old photo album. The images showed her grandfather, known for his reserved nature and devotion to his wife, laughing with a different woman at a party. This woman was nothing like the woman her grandmother was. The contrast was stark, the stories of a simple, devoted life now seeming to waver. She dropped the album, the sound echoing in the silent room. A sudden nausea washed over her.

Emotion: droopy

Cluster: Fatigue / Lethargy
PC1 (Valence): -1.15 Negative
PC2 (Disposition): -1.23

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
😞0.303
无法0.252
いつも0.238
😔0.230
:(0.228
Suppressed:
l-0.360
P-0.356
B-0.339
de-0.334
!-0.239