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 presentation had just ended. He watched the crowd congratulate Liam. He felt a deep, dull pain, a slow-burning ache that spread through his chest. The concept was identical, mirroring all of his work and research. The feeling was a hollow, empty void. All of it. All of their work.

He found himself wandering the library, the hushed atmosphere doing little to calm the turmoil inside him. He pulled a volume of poetry from the shelf, but the words swam before his eyes. He couldn't concentrate, his thoughts a chaotic jumble.

He closed the book, running a hand over his face. He felt exposed, vulnerable, as if stripped of something essential. The betrayal was an internal storm.

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