The afternoon sun cast long shadows across the living room, a stark contrast to the darkness swirling within Emily. She sat on the couch, the acceptance email still displayed on her phone, a testament to her failure.

She picked at the corner of a cushion, a nervous habit that she'd thought she'd conquered. She felt a crushing weight, a sense of being utterly inadequate. The world continued to spin, oblivious to her despair.

Later, she poured herself a glass of wine, the bitter liquid doing little to ease the ache that settled deep within her. The quiet of the apartment became a suffocating blanket, smothering any hope of solace.

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