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 weight of the past seemed to press down on her. The house, once a haven, now felt like a prison. Every room held a memory, a whisper of a life she no longer had. The silence of the house was oppressive, broken only by the rhythmic thumping of her own heartbeat.

She found herself replaying conversations in her head, agonizing over every word, every gesture. Sleep eluded her, replaced by a constant state of restless exhaustion. She stared at the ceiling, the darkness mirroring the emptiness within.

One evening, her father found her sitting on the porch, staring out at the stars. He sat beside her, his silence a comfort. She leaned her head against his shoulder, a silent testament to her pain.

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