CHATS-lab

All

5 repositories

ToolShield
Public
Official implementation for paper "Unsafer in Many Turns: Benchmarking and Defending Multi-Turn Safety Risks in Tool-Using Agents"
Python
•
MIT License
•0•11•0•0•Updated Feb 24, 2026Feb 24, 2026
verbalized-sampling
Public
Verbalized Sampling, a training-free prompting strategy to mitigate mode collapse in LLMs by requesting responses with probabilities. Achieves 2-3x diversity im…
python diversity gemini
python diversity gemini sampling gpt creative-writing claude synthetic-data mode-collapse large-language-models
Python
•
Other
•82•706•0•0•Updated Jan 3, 2026Jan 3, 2026
LLMs_Encode_Harmfulness_Refusal_Separately
Public
Python
•1•29•1•0•Updated Dec 14, 2025Dec 14, 2025
persuasive_jailbreaker
Public
Persuasive Jailbreaker: we can persuade LLMs to jailbreak them!
HTML
•
Apache License 2.0
•29•351•1•0•Updated Oct 17, 2025Oct 17, 2025
KokoMind
Public
KokoMind: Can LLMs Understand Social Interactions?
nlp deep-learning neural-network
nlp deep-learning neural-network language-model gpt-4 chatgpt
JavaScript
•
Apache License 2.0
•8•104•3•0•Updated Oct 3, 2023Oct 3, 2023