Yibo Hu Research Group

Popular repositories Loading

1. Easier-to-Mislead-Than-to-Correct Easier-to-Mislead-Than-to-Correct Public

   Code and data for “Easier to Mislead Than to Correct”: LLM conformity, social influence, and harmful/beneficial revision.

   Python 3

2. DoseBench DoseBench Public

   Benchmark and evaluation framework for reliability analysis of OTC medication reasoning in LLMs.

3. event-coding-reliability event-coding-reliability Public

   Python

4. llm-speaker-free-floor llm-speaker-free-floor Public

   Code and reproduction package for 'Most LLM Conformity Needs No Speaker: Measuring the Speaker-Free Floor in Peer-Pressure Benchmarks' (Hu & Qu, 2026).

5. observability-boundary observability-boundary Public

   Code and reproduction package for 'When Local Monitors Miss Compositional Harm: Diagnosing Distributed Backdoors in Multi-Agent Systems'

6. safety-flag-benchmark safety-flag-benchmark Public

   Python