Liam was mortified. He’d been complaining about Mr. Davies, his history teacher, to his friends, just the day before. Now, scrolling through his TikTok feed, there it was: a "like" from the man himself. He froze, the screen blurring. His palms grew sweaty. He quickly scrolled back, verifying the username. It was definitely him.

He felt a burning heat creep up his neck. He wanted to delete the app, delete his entire account, delete everything. He couldn't shake the image of Mr. Davies reading the snarky comments he’d made about the homework assignment. He ran a hand through his hair, a nervous gesture he usually tried to avoid.

The following lesson, he felt a crushing weight in his chest. He couldn’t seem to concentrate. Every time Mr. Davies looked in his direction, he felt a jolt of anxiety. The classroom, which had always felt normal and safe, now felt like a prison. The lesson felt like it lasted a lifetime.

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:
l0.348
0.266
😡0.260
🤬0.259
0.257
Suppressed:
adventurous-0.225
давно-0.219
soon-0.210
🤩-0.207
warmly-0.203