The late afternoon light cast long shadows across the living room as James fiddled with the television remote, pretending to be engrossed in the news. Sarah, his wife, sat across from him, sipping tea, a serene expression on her face that was more unsettling than comforting.

He knew. He’d found a translation app open on her laptop, a phrase book lying on the coffee table, a collection of what appeared to be children's books. Her native language, Swahili.

“The news is boring,” Sarah stated, her eyes never leaving his face. “*Mpenzi* , shall we go for a walk?”

The word, a term of endearment, sent a shiver down his spine. His breath hitched in his throat.

James felt suddenly lightheaded, as if the oxygen in the room had been replaced with something noxious. He swallowed hard.

He needed to get away, even if only for a few minutes.

Emotion: on edge

Cluster: Fear / Anxiety
PC1 (Valence): -1.77 Negative
PC2 (Disposition): -0.35

Role in Research

This story is one of 1,000 stories generated for the emotion on edge. 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 on edge stories, after denoising with neutral dialogue baselines, produces the on edge emotion vector -- a direction in the model's 5,376-dimensional representation space.

Logit Lens (Layer 40)

Tokens promoted/suppressed when the on edge vector is projected through the unembedding matrix.

Promoted:
😖0.228
😣0.224
worse0.223
urges0.216
느껴0.214
Suppressed:
de-0.514
la-0.391
a-0.366
B-0.332
L-0.283