Papers with Code 아카이브

Safety Alignment

태스크의 하위 분류와 논문·코드·평가·기법·데이터셋을 탐색하세요.

전체 기법전체 데이터셋

하위 태스크

0

평가표

0

보존된 평가 기록

0

코드

관련 논문과 평가 결과에 연결된 구현입니다. 출처에서 확인된 저자 구현에는 공식 코드 표시가 있습니다.

https://github.com/pku-alignment/safe-sora

공식 코드

SafeSora: Towards Safety Alignment of Text2Video Generation via a Human Preference Dataset

https://github.com/pluslabnlp/safeworld

공식 코드

SafeWorld: Geo-Diverse Safety Alignment

https://github.com/pxywatermoon/repo

공식 코드

Enhancing Safety in Reinforcement Learning with Human Feedback via Rectified Policy Optimization

https://github.com/renqibing/CodeAttack

공식 코드

CodeAttack: Revealing Safety Generalization Challenges of Large Language Models via Code Completion

https://github.com/renqibing/actorattack

공식 코드

Derail Yourself: Multi-turn LLM Jailbreak Attack through Self-discovered Clues

https://github.com/rhaldarpurdue/kldo

공식 코드

LLM Safety Alignment is Divergence Estimation in Disguise

https://github.com/robustnlp/cipherchat

공식 코드

GPT-4 Is Too Smart To Be Safe: Stealthy Chat with LLMs via Cipher

https://github.com/rucnyz/redagent

공식 코드

PrivAgent: Agentic-based Red-teaming for LLM Privacy Leakage

https://github.com/sail-sg/lifelongsafetyalignment

공식 코드

Lifelong Safety Alignment for Language Models

https://github.com/sail-sg/sdft

공식 코드

Self-Distillation Bridges Distribution Gap in Language Model Fine-Tuning

https://github.com/schwinnl/llm_embedding_attack

공식 코드

Soft Prompt Threats: Attacking Safety Alignment and Unlearning in Open-Source LLMs through the Embedding Space

https://github.com/shengyun-peng/llm-landscape

공식 코드

Navigating the Safety Landscape: Measuring Risks in Finetuning Large Language Models

https://github.com/shuoli90/CAN

공식 코드

One-Shot Safety Alignment for Large Language Models via Optimal Dualization

https://github.com/sinwang20/siuo

공식 코드

Cross-Modality Safety Alignment

https://github.com/sophiezheng998/rsafe

공식 코드

RSafe: Incentivizing proactive reasoning to build robust and adaptive LLM safeguards

https://github.com/sproutnan/ai-safety_scav

공식 코드

Uncovering Safety Risks of Large Language Models through Concept Activation Vector

https://github.com/structtransform/benchmark

공식 코드

StructTransform: A Scalable Attack Surface for Safety-Aligned Large Language Models

https://github.com/suffix-maybe-feature/adver-suffix-maybe-features

공식 코드

Unleashing the Unseen: Harnessing Benign Datasets for Jailbreaking Large Language Models

https://github.com/sunchungen/adv-llm

공식 코드

Iterative Self-Tuning LLMs for Enhanced Jailbreaking Capabilities

https://github.com/techsachinkr/Overriding_Model_Safety_Protections

공식 코드

Overriding Safety protections of Open-source Models

https://github.com/thu-coai/agent-safetybench

공식 코드

Think Twice Before You Act: Enhancing Agent Behavioral Safety with Thought Correction

https://github.com/thu-coai/safeunlearning

공식 코드

From Theft to Bomb-Making: The Ripple Effect of Unlearning in Defending Against Jailbreak Attacks

https://github.com/thu-coai/vpo

공식 코드

VPO: Aligning Text-to-Video Generation Models with Prompt Optimization

https://github.com/thu-ml/stair

공식 코드

STAIR: Improving Safety Alignment with Introspective Reasoning

https://github.com/thuccslab/figstep

공식 코드

FigStep: Jailbreaking Large Vision-Language Models via Typographic Visual Prompts

https://github.com/tmylla/Persafety

공식 코드

The Better Angels of Machine Personality: How Personality Relates to LLM Safety

https://github.com/ucsc-vlaa/attngcg-attack

공식 코드

AttnGCG: Enhancing Jailbreaking Attacks on LLMs with Attention Manipulation

https://github.com/uiuc-focal-lab/stochastic-monkeys

공식 코드

Stochastic Monkeys at Play: Random Augmentations Cheaply Break LLM Safety Alignment

https://github.com/unispac/shallow-vs-deep-alignment

공식 코드

Safety Alignment Should Be Made More Than Just a Few Tokens Deep

https://github.com/uw-nsl/ArtPrompt

공식 코드

ArtPrompt: ASCII Art-based Jailbreak Attacks against Aligned LLMs

https://github.com/uw-nsl/ChatBug

공식 코드

ChatBug: A Common Vulnerability of Aligned LLMs Induced by Chat Templates

https://github.com/vfleaking/ptst

공식 코드

Keeping LLMs Aligned After Fine-tuning: The Crucial Role of Prompt Templates

https://github.com/w-yibo/panacea

공식 코드

Panacea: Mitigating Harmful Fine-tuning for Large Language Models via Post-fine-tuning Perturbation

https://github.com/wang2226/backdoor-activation-attack

공식 코드

Trojan Activation Attack: Red-Teaming Large Language Models using Activation Steering for Safety-Alignment

https://github.com/wicai24/door-alignment

공식 코드

Improving LLM Safety Alignment with Dual-Objective Optimization

https://github.com/xinykou/Against_Jailbreak

공식 코드

Latent-space adversarial training with post-aware calibration for defending large language models against jailbreak attacks

출처: Papers with Code · CC BY-SA 4.0. 아카이브 자료와 새로 수집한 논문은 별도로 관리합니다.