Pinned Loading

1. emergent-misaligned-agents emergent-misaligned-agents Public

   In this repository, we explore the notion of emergent misalignment in the context of tool-augmented large language models. Tested models are finetuned on partially incorrect datasets to induce misa…

   Python

2. Protocol66 Protocol66 Public

   Protocol 66 is an automated red‑teaming harness for AI agents. It inspects an agent’s configuration, synthesizes adversarial prompts tailored to its tools and permissions, simulates responses via C…

   TypeScript

3. dorukersoy47/lighthouse-ai dorukersoy47/lighthouse-ai Public

   Python

4. multi-step-agent-rl-infra multi-step-agent-rl-infra Public

   RL training infrastructure for multi-step web agents. Generates tasks at controllable difficulty (planning horizon), validates with oracle, evaluates efficiency-accuracy tradeoff

   Python