I'm Junxiao Yang, a first-year Phd student from Tsinghua University.
π Iβm currently working on LLM safety, Reliability and Alignment.
π± Here's my homepage: https://yangjunx21.github.io/
I'm Junxiao Yang, a first-year Phd student from Tsinghua University.
π Iβm currently working on LLM safety, Reliability and Alignment.
π± Here's my homepage: https://yangjunx21.github.io/
π« CAMEL: The first and the best multi-agent framework. Finding the Scaling Law of Agents. https://www.camel-ai.org
Open-source AI agent desktop app for Windows & macOS. One-click install Claude Code, MCP tools, and Skills β with sandbox isolation, multi-model support, and Feishu/Slack integration.
AISafetyLab: A comprehensive framework covering safety attack, defense, evaluation and paper list.
π Loong: Synthesize Long CoTs at Scale through Verifiers.
[ACL 2025] Guiding not Forcing: Enhancing the Transferability of Jailbreaking Attacks on LLMs via Removing Superfluous Constraints
Open-source Claude Design alternative. One-click import your Claude Code / Codex API key. Prompt β prototype / slides / PDF. Multi-model (Claude, GPT, Gemini, Kimi, GLM, Ollama). BYOK, local-first,β¦