GLM-OCR: Accurate × Fast × Comprehensive

The official implementation of the paper "AgentDyn: A Dynamic Open-Ended Benchmark for Evaluating Prompt Injection Attacks of Real-World Agent Security System".

Nano vLLM

Contexts Optical Compression

[NeurIPS 2023] MeZO: Fine-Tuning Language Models with Just Forward Passes. https://arxiv.org/abs/2305.17333

A benchmark for evaluating the robustness of LLMs and defenses to indirect prompt injection attacks.

The Granite Guardian models are designed to detect risks in prompts and responses.

A benchmark for prompt injection detection systems.

Provide with pre-build flash-attention package wheels on Linux and Windows platforms using GitHub Actions

Official codebase for "STAIR: Improving Safety Alignment with Introspective Reasoning"

A list of recent papers about adversarial learning

slime is an LLM post-training framework for RL Scaling.

Agent Reinforcement Trainer: train multi-step agents for real-world tasks using GRPO. Give your agents on-the-job training. Reinforcement learning for Qwen2.5, Qwen3, Llama, and more!

My learning notes for ML SYS.

A project to improve skills of large language models

Official implementation of the WASP web agent security benchmark

Official Implementation of implicit reference attack

Emoji Attack [ICML 2025]

dLLM: Simple Diffusion Language Modeling

A Python library for guardrail models evaluation.

Open One-Stop Moderation Tools for Safety Risks, Jailbreaks, and Refusals of LLMs

NeMo Guardrails is an open-source toolkit for easily adding programmable guardrails to LLM-based conversational systems.

Code for the paper "Defeating Prompt Injections by Design"

Patch Linux executables for compatibility with older glibc

Patch Linux executables for compatibility with older glibc

🪢 Open source LLM engineering platform: LLM Observability, metrics, evals, prompt management, playground, datasets. Integrates with OpenTelemetry, Langchain, OpenAI SDK, LiteLLM, and more. 🍊YC W23

Internal Consistency Regularization (CROW) for LLM Backdoor Elimination - Paper accepted to ICML 2025

Paper Link- https://arxiv.org/abs/2510.21910