He’d always thought he and Liam had a bond. They’d spent hours building Lego fortresses, attending baseball games, sharing inside jokes only they understood. Now, sifting through the evidence – the secret Twitter account, the coded messages – a sickness churned in his stomach. A tight band constricted his chest, making it difficult to breathe.

He slammed his fist against the kitchen table, the sudden noise shattering the fragile quiet. The chipped paint seemed to mirror his own flaking emotions. Every shared memory, every father-son moment, felt tainted, like it had all been a carefully constructed performance.

He wanted to scream, to lash out, but all that came was a choked sob. He retreated to the couch, clutching the worn cushion. He felt utterly adrift, a ship lost at sea without a sail.

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
πŸ˜”0.359
haunting0.349
😒0.342
Suppressed:
de-0.503
la-0.455
B-0.352
πŸ€“-0.332
!-0.299