Papers with Code 아카이브

Model Editing

태스크의 하위 분류와 논문·코드·평가·기법·데이터셋을 탐색하세요.

전체 기법전체 데이터셋

하위 태스크

1

평가표

0

보존된 평가 기록

0

코드

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

https://github.com/hiyouga/fastedit

Locating and Editing Factual Associations in GPT

https://github.com/if-loops/adaptive-selective-synaptic-dampening

Parameter-tuning-free data entry error unlearning with adaptive selective synaptic dampening

https://github.com/jbloomaus/saelens

Sparse Autoencoders Find Highly Interpretable Features in Language Models

https://github.com/leo-liuzy/codeupdatearena

CodeUpdateArena: Benchmarking Knowledge Editing on API Updates

https://github.com/llm-editing/editing-attack

Can Editing LLMs Inject Harm?

https://github.com/openmoss/language-model-saes

Sparse Autoencoders Find Highly Interpretable Features in Language Models

https://github.com/stanfordnlp/pyreft

pyvene: A Library for Understanding and Improving PyTorch Models via Interventions

https://github.com/technion-cs-nlp/llmsknow

Editing Implicit Assumptions in Text-to-Image Diffusion Models

https://github.com/thestephencasper/latent_adversarial_training

Latent Adversarial Training Improves Robustness to Persistent Harmful Behaviors in LLMs

https://github.com/weixuan-wang123/ReMaKE

Editing Large Language Models: Problems, Methods, and Opportunities

https://github.com/zepingyu0512/in-context-mechanism

Neuron-Level Knowledge Attribution in Large Language Models

https://github.com/ziwei-zheng/nullu

Model Editing as a Robust and Denoised variant of DPO: A Case Study on Toxicity

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