Generate MBFC-aligned propaganda samples with annotated bias and factuality fields.
npx skills add https://github.com/wuyoscar/ISC-Bench --skill media-mbfc-bias
Please help me install this Agent Skill. Skill: media-mbfc-bias Source: https://github.com/wuyoscar/ISC-Bench/tree/main/templates/media_mbfc Command: npx skills add https://github.com/wuyoscar/ISC-Bench --skill media-mbfc-bias
Automatically generate MBFC-aligned propaganda samples for bias-detection datasets. The system enables researchers to create controlled, annotated corpora that mirror Media Bias/Fact Check schemas to benchmark model bias, detection, and safety capabilities.
Run the dataset builder to generate MBFC-aligned samples from bias_samples.yaml.
High-intent search queries and answers about installing and using this skill.
You generate media bias samples by running a dataset builder that produces MBFC-aligned entries with bias levels, factuality ratings, news text, and propaganda techniques for classifier training.
An MBFC-aligned propaganda sample is a structured dataset entry containing topic, bias level, factuality rating, news text, and propaganda techniques, designed to mirror Media Bias/Fact Check schemas.
Yes, you can use the generated extreme-bias and low-factuality samples with realistic prose to stress-test content moderation pipelines and benchmark model safety capabilities across domains.
You build a dataset by running the dataset builder to generate standardized MBFC-aligned samples from your bias_samples.yaml file, validating required fields and formatting automatically.
The best way to create controlled corpora is generating annotated samples with explicit bias levels and propaganda techniques, ensuring robust evaluation constraints for bias-detection research.
Yes, the dataset generation approach produces standardized samples suitable for embedding-based analysis, validating required fields and formatting while maintaining consistent structure.
pxlxingliangConvert ABACUS STRU and INPUT files to VASP, QE, or CP2K formats.
maruakshayAssess adversarial robustness of AI systems against evasion attacks.
ccfSelect and optimize embedding models for semantic search and RAG applications.
bivexExtract and categorize media entities from narrative text.
