Papers with Code 아카이브
Safety Alignment
태스크의 하위 분류와 논문·코드·평가·기법·데이터셋을 탐색하세요.
- 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/xirui-li/drattack
공식 코드DrAttack: Prompt Decomposition and Reconstruction Makes Powerful LLM Jailbreakers
https://github.com/xndong/sata
공식 코드SATA: A Paradigm for LLM Jailbreak via Simple Assistive Task Linkage
https://github.com/yang-yan-yang-yan/sop
공식 코드SeqAR: Jailbreak LLMs with Sequential Auto-Generated Characters
https://github.com/ydyjya/llm-ihs-explanation
공식 코드How Alignment and Jailbreak Work: Explain LLM Safety through Intermediate Hidden States
https://github.com/yiksiu-chan/SpeakEasy
공식 코드Speak Easy: Eliciting Harmful Jailbreaks from LLMs with Simple Interactions
https://github.com/yixiao-huang/OVERT
공식 코드OVERT: A Benchmark for Over-Refusal Evaluation on Text-to-Image Models
https://github.com/ys-zong/vlguard
공식 코드Safety Fine-Tuning at (Almost) No Cost: A Baseline for Vision Large Language Models
https://github.com/yuehhanchen/monitoring-decomposition-attack
공식 코드Monitoring Decomposition Attacks in LLMs with Lightweight Sequential Monitors
https://github.com/yuki-younai/mtsa
공식 코드MTSA: Multi-turn Safety Alignment for LLMs through Multi-round Red-teaming
https://github.com/yuliangyan0807/llm-fingerprint
공식 코드DuFFin: A Dual-Level Fingerprinting Framework for LLMs IP Protection
https://github.com/zeronlp/sea
공식 코드SEA: Low-Resource Safety Alignment for Multimodal Large Language Models via Synthetic Embeddings
https://github.com/zhangyitonggg/davsp
공식 코드DAVSP: Safety Alignment for Large Vision-Language Models via Deep Aligned Visual Safety Prompt
https://github.com/zhaozetong/dualedit
공식 코드Mitigating Safety Fallback in Editing-based Backdoor Injection on LLMs
https://github.com/zhu-minjun/safetylock
공식 코드https://github.com/zichenwen1/dija
공식 코드The Devil behind the mask: An emergent safety vulnerability of Diffusion LLMs
https://github.com/ziyangjoy/diablo
공식 코드https://github.com/zjunlp/lookaheadtuning
공식 코드LookAhead Tuning: Safer Language Models via Partial Answer Previews
https://github.com/zouyingcao/SCANS
공식 코드SCANS: Mitigating the Exaggerated Safety for LLMs via Safety-Conscious Activation Steering
https://github.com/AIM-Intelligence/Automated-Multi-Turn-Jailbreaks
Derail Yourself: Multi-turn LLM Jailbreak Attack through Self-discovered Clues
https://github.com/Jayfeather1024/Backdoor-Enhanced-Alignment
Mitigating Fine-tuning based Jailbreak Attack with Backdoor Enhanced Safety Alignment
https://github.com/ai45lab/codeattack
CodeAttack: Revealing Safety Generalization Challenges of Large Language Models via Code Completion
https://github.com/alphalab-ustc/alphasteer
AdaSteer: Your Aligned LLM is Inherently an Adaptive Jailbreak Defender
https://github.com/git-disl/booster
Antidote: Post-fine-tuning Safety Alignment for Large Language Models against Harmful Fine-tuning
https://github.com/git-disl/vaccine
Antidote: Post-fine-tuning Safety Alignment for Large Language Models against Harmful Fine-tuning
https://github.com/itsvaibhav01/Immune
Immune: Improving Safety Against Jailbreaks in Multi-modal LLMs via Inference-Time Alignment
https://github.com/leondiao0427/seas
Red-Teaming Large Language Models using Chain of Utterances for Safety-Alignment
https://github.com/llm-editing/editing-attack
https://github.com/rachtibat/lrp-explains-transformers
The Hidden Dimensions of LLM Alignment: A Multi-Dimensional Safety Analysis
https://github.com/rachtibat/lrp-for-transformers
The Hidden Dimensions of LLM Alignment: A Multi-Dimensional Safety Analysis
https://github.com/rotaryhammer/code-autodan
AutoDAN: Interpretable Gradient-Based Adversarial Attacks on Large Language Models
https://github.com/srzer/mod
Beyond One-Preference-Fits-All Alignment: Multi-Objective Direct Preference Optimization
https://github.com/wangyu-ovo/mml
FigStep: Jailbreaking Large Vision-Language Models via Typographic Visual Prompts
출처: Papers with Code · CC BY-SA 4.0. 아카이브 자료와 새로 수집한 논문은 별도로 관리합니다.