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 scent of freshly cut wood assaulted her nostrils as she watered her lawn. The fence posts stood like soldiers, rigid and unyielding, a testament to finality. The hammering had ceased, replaced by an unsettling stillness. Across the now-divided yard, she saw Mrs. Davis tending her prize-winning petunias, a bright and cheerful activity that grated on her nerves.

A sigh escaped her lips as she thought about the conversations she used to share with David over the fence. The casual chats, the shared complaints about the weather, the easy camaraderie of neighbors. Those days were gone. She felt a deep, pervasive weariness settling over her.

She lowered the hose, the water now soaking her jeans, as the sharp, clean scent of the wood filled her lungs.

Emotion: heartbroken

Cluster: Sadness / Despair
PC1 (Valence): -2.60 Negative
PC2 (Disposition): -1.10

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
😞0.433
S0.385
😔0.359
haunting0.349
😢0.342
Suppressed:
de-0.503
la-0.455
B-0.352
🤓-0.332
!-0.299