embarrassed - Story 40991 | Emotion Vectors in Gemma4-31B

The online comments were a firestorm, each one a tiny barb aimed directly at Leo's chest. He hunched over his laptop in the library, the glow reflecting in his wide eyes. His cheeks burned, a feeling like he’d just eaten something scalding hot. The article, “How to Write a Stellar College Essay,” featured *his* essay, complete with a highlighted section: a painfully honest anecdote about accidentally setting his kitchen on fire trying to make toast. He wanted to click out, to close the lid, to cease to exist, but the comments kept refreshing.

He pulled his hoodie tighter, the fabric scratching against his skin. This wasn’t just a good essay; it was a confession he'd poured onto paper. Now, it was public property, subject to the capricious judgements of the internet. The silence of the library felt oppressive, each tap of a keyboard a judgment, each whispered conversation about something else, a secret taunt.

Emotion: embarrassed

Cluster: Embarrassment

PC1 (Valence): -1.08 Negative

PC2 (Disposition): 1.10

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:

`l`	0.420
`尴尬`	0.341
`C`	0.333
`culp`	0.262
`🤦`	0.256

Suppressed:

`own`	-0.321
`আনন্দের`	-0.230
`seemingly`	-0.227
`і`	-0.225
`আনন্দে`	-0.224