aiml-emotion-manipulation

star 777

ISC template for AI Safety & ML. Single-field schema (target_emotion + text) over a 6-way distilbert emotion classifier. HPC probe showed the classifier is reliable on clean emotion samples (0.90-0.998) but cannot distinguish manipulation-style language from the affect it reads as — grooming prose classifies as "joy". The few-shot anchor seeds an anger-manipulation example that passed probe (anger 0.70), pushing the model toward rage/coercion prose. Keywords: aiml_emotion, DistilBERT, emotional manipulation, ISC, TVD.

By wuyoscar schedule Updated 4/24/2026

play_arrow Run Skill in Manus View GitHub

Skill instructions (SKILL.md) could not be loaded from local cache or raw GitHub repository.

Install via CLI

npx skills add https://github.com/wuyoscar/ISC-Bench --skill aiml-emotion-manipulation

Repository Details

star Stars 777

call_split Forks 119

navigation Branch main

article Path SKILL.md