The chipped porcelain mug felt heavy in Clara's trembling hands. Steam curled upwards, carrying the ghost of the chamomile tea she’d brewed hours ago, now lukewarm and forgotten. The real estate agent's words echoed in her mind: “Charming Victorian, full of character…though, a bit of history to it.” A history that turned out to be a gruesome unsolved murder. Now, the walls felt like they were closing in, mimicking the way her chest constricted with each shallow breath. She imagined the previous owner, also a woman, also alone. Had she felt this hollow ache, this overwhelming sense of being adrift, in the same rooms? Clara clutched the mug tighter, a desperate attempt to find some semblance of stability.

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