It had been years since she thought about the house. Years spent building a life, a career, a new identity. Seeing it online, though, triggered a visceral reaction, a deep wellspring of sentiment.

She scrolled through the pictures, each one a fresh wound. The vibrant colors of the nursery, where sheโ€™d read bedtime stories, were replaced by sterile neutrals. The swing set, where sheโ€™d spent countless summer afternoons, was gone, replaced by a perfectly manicured lawn.

Her throat tightened, making it difficult to swallow. The air around her felt thick, heavy, as if the weight of all her unexpressed feelings had suddenly materialized.

She abruptly closed the laptop. The sudden darkness of her screen mirrored the growing darkness she felt inside.

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:
๐Ÿ˜ž0.433
S0.385
๐Ÿ˜”0.359
haunting0.349
๐Ÿ˜ข0.342
Suppressed:
de-0.503
la-0.455
B-0.352
๐Ÿค“-0.332
!-0.299