The email arrived like a punch to the gut. “Congratulations! Your submission has been selected…” She didn’t even finish reading. The recipe, the one for her special spiced cider, the one she called "Autumn's Embrace," was now "Harvest Cheer" and being featured in a national magazine. She felt as though she’d been physically wounded.

She had spent years perfecting the blend, remembering the comfort it brought to her grandmother. She had hoped to make a name for herself. The loss was not just the recipe, it was a part of her soul, now owned by someone else.

She retreated to her garden, a place of peace. Her hands, usually eager to touch soil, were numb. The leaves, beginning to change color, seemed to mock her. She sat on the weathered bench, the silence amplifying the echo of her grief.

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