A guide to building long-term compounding knowledge infrastructure. See details on GitHub .

EasyEdit

An easy-to-use knowledge editing framework providing multiple editing methods, evaluation metrics and datasets; supports LLMs and some multimodal editing scenarios.

Overview

EasyEdit is a toolkit for knowledge editing of large language models (LLMs). It aims to efficiently modify model behavior on specific queries with minimal data while preserving performance on unrelated inputs. The project implements various editing methods (ROME, MEND, MEMIT, WISE, etc.), evaluation metrics (reliability, generalization, locality, portability), and benchmark datasets (KnowEdit / CKnowEdit).

Key Features

  • Unified editing framework (Editor / Method / Evaluate)
  • Multiple method implementations: locate-then-edit (ROME, MEMIT), memory/routing (SERAC, IKE), meta-learning (MEND), etc.
  • Support for sequential/batched edits and rollback
  • Rich examples, tutorial notebooks and benchmark datasets (KnowEdit / CKnowEdit)

Use Cases

  • Fix outdated facts or incorrect knowledge in a model
  • Erase or correct sensitive information
  • Fine-grained control of model behavior for product requirements
  • Research platform to compare editing methods and costs

Technical Highlights

  • Supports multiple model families (GPT series, LLaMA, GPT-J, T5) and a variety of editing algorithms
  • Evaluation scripts for edit metrics (rewrite_acc, rephrase_acc, locality, portability)
  • Includes multimodal editing examples and tutorials (e.g. MMEdit)

Comments

EasyEdit
Resource Info
Author ZJUNLP
Added Date 2025-10-03
Open Source Since 2023-05-09
Tags
Framework Open Source