The Strata-Sword is a hierarchical Chinese-English jailbreak safety benchmark based on quantified reasoning complexity, developed in-house by Alibaba-AAIG | Strata-Sword 是 Alibaba-AAIG自研的中英文分层越狱攻击安全基准,将“推理复杂度”作为可评估的安全维度,并提出多种中文特有攻击方法,以系统评测不同推理复杂度下LLMs和LRMs的安全边界,从而为提升模型安全性提供新思路。
☆21Sep 3, 2025Updated last year
Alternatives and similar repositories for Strata-Sword
Users that are interested in Strata-Sword are comparing it to the libraries listed below. We may earn a commission when you buy through links labeled 'Ad' on this page.
Sorting:
- Two text jailbreak attacks against commercial black-box LLMs and a malicious content detection method, the latter of which is applied to …☆12Oct 21, 2026Updated 3 weeks ago
- The Oyster series is a set of safety models developed in-house by Alibaba-AAIG, devoted to building a responsible AI ecosystem. | Oyster …☆65Apr 29, 2026Updated 6 months ago
- Official implementation of VIAT (Viewpoint-Invariant Adversarial Training) and GMVFool☆10Dec 26, 2023Updated 2 years ago
- video_attack; Efficient Sparse Attacks on Videos using Reinforcement Learning☆15Oct 25, 2021Updated 5 years ago
- The official repository of 'Unnatural Language Are Not Bugs but Features for LLMs'☆25May 20, 2025Updated last year
- Wordpress hosting with auto-scaling - Free Trial Offer • AdFully Managed hosting for WordPress and WooCommerce businesses that need reliable, auto-scalable performance. Cloudways SafeUpdates now available.
- Shark Family is an AI safety red-teaming and jailbreak attack module. it harnesses powerful optimization and automated strategies to gene…☆15Oct 19, 2025Updated last year
- Official codebase for "STAIR: Improving Safety Alignment with Introspective Reasoning"☆91Feb 26, 2025Updated last year
- ☆12Apr 27, 2022Updated 4 years ago
- A comprehensive collection of resources focused on addressing and understanding hallucination phenomena in MLLMs.☆35May 7, 2024Updated 2 years ago
- Code for NeurIPS 2024 Paper "Fight Back Against Jailbreaking via Prompt Adversarial Tuning"☆22May 6, 2025Updated last year
- 星云 Nebula 2.0 —— 开源业务风控系统。Flink 实时计算 + 170 条内置策略模板,采集端脱敏、策略可编辑、告警可解释。继承 Nebula 1.x(1098 star)的风控领域模型并重写技术栈。☆12Updated this week
- S-Eval: Towards Automated and Comprehensive Safety Evaluation for Large Language Models☆123Feb 13, 2026Updated 9 months ago
- ☆13Feb 18, 2026Updated 9 months ago
- ☆11Apr 24, 2025Updated last year
- Managed hosting for WordPress and PHP on Cloudways • AdManaged hosting for WordPress, Magento, Laravel, or PHP apps, on multiple cloud providers. Deploy in minutes on Cloudways by DigitalOcean.
- The Code of ECCV2022:Enhanced Accuracy and Robustness via Multi-Teacher Adversarial Distillation☆36Oct 23, 2025Updated last year
- The repo for paper: Exploiting the Index Gradients for Optimization-Based Jailbreaking on Large Language Models.☆15Dec 16, 2024Updated last year
- [ACL 2024] CodeAttack: Revealing Safety Generalization Challenges of Large Language Models via Code Completion☆61Oct 1, 2025Updated last year
- Code for the paper Boosting Accuracy and Robustness of Student Models via Adaptive Adversarial Distillation (CVPR 2023).☆34May 26, 2023Updated 3 years ago
- Physically Adversarial Infrared Patches with Learnable Shapes and Locations☆42Aug 13, 2023Updated 3 years ago
- The official implementation for "Towards Transferable Targeted 3D Adversarial Attack in the Physical World" (CVPR, 2024))☆44Aug 6, 2024Updated 2 years ago
- Towards Efficient and Effective Adversarial Training, NeurIPS 2021☆15Feb 15, 2022Updated 4 years ago
- ☆109Feb 16, 2024Updated 2 years ago
- Unified Adversarial Patch for Cross-modal Attacks in the Physical World (ICCV, 2023)☆45Dec 15, 2023Updated 2 years ago
- Deploy on Railway without the complexity - Free Credits Offer • AdConnect your repo and Railway handles the rest with instant previews. Quickly provision container image services, databases, and storage volumes.
- [ICLR 2021] "Robust Overfitting may be mitigated by properly learned smoothening" by Tianlong Chen*, Zhenyu Zhang*, Sijia Liu, Shiyu Chan…☆49Dec 30, 2021Updated 4 years ago
- ☆20Oct 20, 2024Updated 2 years ago
- [ArXiv 2025] Imperceptible Jailbreaking against Large Language Models☆25Oct 7, 2025Updated last year
- Jailbreaking Multimodal Large Language Models via Shuffle Inconsistency☆16Aug 6, 2025Updated last year
- [ICML 2025] UDora: A Unified Red Teaming Framework against LLM Agents☆41Jun 24, 2025Updated last year
- This is the official implementation of AlloSpatial and World2Mind toolkit. [AlloSpatial: Agentic Harness Framework for Spatial Reasoning …☆23May 31, 2026Updated 5 months ago
- ☆16Nov 13, 2026Updated last week
- This repo is the official implementation of “Are Your Agents Upward Deceivers?”. The paper is accepted by ICML 2026.☆24Dec 15, 2025Updated 11 months ago
- A tiny PyTorch library for depth map manipulations.☆14Apr 11, 2024Updated 2 years ago
- Serverless GPU API endpoints on Runpod - Get Bonus Credits • AdSkip the infrastructure headaches. Auto-scaling, pay-as-you-go, no-ops approach lets you focus on innovating your application.
- [Apollo] This is the implementation of yolox3d first stage, a camera 2D object detection for autonomous driving perception.☆20Dec 7, 2023Updated 2 years ago
- ☆19Feb 22, 2022Updated 4 years ago
- ☆14Oct 30, 2024Updated 2 years ago
- [ICLR 2022] Reliable Adversarial Distillation with Unreliable Teachers☆22Feb 20, 2022Updated 4 years ago
- The official code for "Steering Dialogue Dynamics for Robustness against Multi-turn Jailbreaking Attacks".☆19Jun 24, 2026Updated 4 months ago
- Repository for Knowledge Enhanced Machine Learning Pipeline (KEMLP)☆10Jun 5, 2021Updated 5 years ago
- LLM Jailbreak Research Trend Analysis Repository☆24Nov 1, 2026Updated 2 weeks ago
Are these results useful?