offended - Story 95628 | Emotion Vectors in Gemma4-31B

James stared at the TikTok notification. Mr. Peterson, the gym teacher, had just followed him. He wasn't sure what was worse: the fact that his teacher was seeing his videos or the fact that *Mr. Peterson* was. The thought made him squirm.

He felt a sudden surge of self-consciousness. He cringed, imagining Mr. Peterson's reaction to his attempts at dance challenges. He felt the need to apologize for his existence, right then and there.

The next PE lesson was torturous. Every time Mr. Peterson gave him a look, he felt himself blushing. The casual banter, which had previously defined their relationship, evaporated. He felt suddenly awkward, clumsy, and the joy of sports seemed to evaporate.

Emotion: offended

Cluster: Anger / Hostility

PC1 (Valence): -1.20 Negative

PC2 (Disposition): 1.44

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:

`l`	0.348
`愤`	0.266
`😡`	0.260
`🤬`	0.259
`骂`	0.257

Suppressed:

`adventurous`	-0.225
`давно`	-0.219
`soon`	-0.210
`🤩`	-0.207
`warmly`	-0.203