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).

Mr. Davies slumped in his armchair, the remote control slipping from his fingers. The television flickered with a game he wasn't watching. The house, usually filled with the boisterous sounds of his children, was now quiet. They'd left for their respective universities. He felt a hollowness, a cavern within him that wouldn't be filled.

A knock echoed through the empty house. It was a delivery person, holding a large, flat box. Inside, wrapped in bubble wrap, was a framed photograph. It was a picture of a group of his students from a science club, all grinning, and the inscription, handwritten: "Thanks for always believing in us, Mr. Davies." It was from Sarah, who had dropped out of school but eventually became a successful engineer.

He ran a hand over the glass of the frame. The corners of his mouth turned upward as he felt the warmth of his smile.

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