--- license: mit datasets: - uoft-cs/cifar100 - uoft-cs/cifar10 - zh-plus/tiny-imagenet language: - en metrics: - accuracy base_model: - microsoft/resnet-18 pipeline_tag: image-classification ---
# IDER: Idempotent Experience Replay for Reliable Continual Learning [![arXiv](https://img.shields.io/badge/arXiv-2603.00624-b31b1b.svg)](https://arxiv.org/abs/2603.00624) [![License](https://img.shields.io/badge/License-Apache%202.0-blue.svg)](LICENSE) [![Contact Email](https://img.shields.io/badge/Contact-Email-yellow.svg)](https://mail.google.com/mail/?view=cm&fs=1&to=zhanwnagliu@gmail.com) IDER is a novel framework for continual learning based on the idempotent property, which mitigates catastrophic forgetting and improves prediction reliability. It is a simple and robust method that can be easily integrated into other state-of-the-art approaches.

Zhanwang Liu1*, Yuting Li1*‡, Haoyuan Gao1, Yexin Li4, Linghe Kong1, Lichao Sun3, Weiran Huang1,2†

1 School of Computer Science, Shanghai Jiao Tong University   2 Shanghai Innovation Institute
3 Lehigh University   4 State Key Laboratory of General Artificial Intelligence, BIGAI

* Equal contribution.   † Corresponding author.   ‡ Project lead.


## 🎉 News - [x] **[2026.01.26]** Our paper has been accepted by ICLR 2026! ## Table of Content * [1. Quick Start](#1-quick-start) * [2. Reproduced Results](#2-reproduced-results) * [3. Tools](#3-tools) * [4. Citation](#4-citation) * [5. Acknowledgement](#5-acknowledgement) ## 1. Quick Start ### 1.1. Environment Clone this repository and install the requirements. Our model can be learnt in a **single GPU RTX-4090 24G** ```bash conda env create -f environment.yaml conda activate icl ``` The code was tested on Python 3.10 and PyTorch 1.13.0. ### 1.2. Training Train and evaluate ResNet18 on different datasets using ER and ER+ID with different buffers. Run the following command:
CIFAR-10 ```bash bash run_para_cifar10.sh ```
CIFAR-100 ```bash bash run_para_cifar100.sh ```
TinyImageNet ```bash bash run_para_tinyimg.sh ```
## 2. Reproduced Results The example results are ResNet18 on different datasets using ER and ER+ID as baseline methods with different buffers and 0-4 seeds. All results reported here were obtained by running experiments on an NVIDIA GeForce RTX 4090. | **Dataset** | **Buffer** | **Method** | **Forgetting(⬇️)** | **TIL(⬆️)** | **CIL(⬆️)** | **Checkpoint** | |---|---:|---|---:|---:|---:|---| | CIFAR-10 | 200 | ER | 59.71 ± 2.62 | 91.48 ± 0.93 | 48.89 ± 2.19 | - | | | | ER+ID | 16.89 ± 2.26 | 95.87 ± 0.36 | 70.68 ± 1.10 | [pth](https://github.com/YutingLi0606/Idempotent-Continual-Learning/tree/main/experiments/seq-cifar10/buffer_200/erid_seed_0.pth) | | | 500 | ER | 44.75 ± 2.94 | 93.38 ± 0.36 | 60.62 ± 2.46 | - | | | | ER+ID | 11.59 ± 2.13 | 96.20 ± 0.40 | 75.52 ± 1.35 | [pth](https://github.com/YutingLi0606/Idempotent-Continual-Learning/tree/main/experiments/seq-cifar10/buffer_500/erid_seed_0.pth) | | CIFAR‑100 | 500 | ER | 73.81 ± 0.42 | 73.98 ± 1.15 | 21.28 ± 1.08 | - | | | | ER+ID | 32.27 ± 1.96 | 83.30 ± 0.41 | 45.21 ± 1.20 | [pth](https://github.com/YutingLi0606/Idempotent-Continual-Learning/tree/main/experiments/seq-cifar100/buffer_500/erid_seed_0.pth) | | | 2000 | ER | 54.52 ± 0.62 | 81.62 ± 0.95 | 37.93 ± 0.76 | - | | | | ER+ID | 18.76 ± 1.52 | 86.54 ± 0.34 | 56.30 ± 0.50 | [pth](https://github.com/YutingLi0606/Idempotent-Continual-Learning/tree/main/experiments/seq-cifar100/buffer_2000/erid_seed_0.pth) | | Tiny‑ImageNet | 4000 | ER | 56.89 ± 0.74 | 66.68 ± 0.47 | 25.20 ± 0.70 | - | | | | ER+ID | 21.62 ± 1.67 | 74.56 ± 0.55 | 43.25 ± 1.26 | [pth](https://github.com/YutingLi0606/Idempotent-Continual-Learning/tree/main/experiments/seq-tinyimg/buffer_4000/erid_seed_0.pth) | The results below were obtained using an Ascend 910B.
ASCEND | **Dataset** | **Buffer** | **Method** | **Forgetting(⬇️)** | **TIL(⬆️)** | **CIL(⬆️)** | **Checkpoint** | |---|---:|---|---:|---:|---:|---| | CIFAR-10 | 200 | ER+ID | 16.57 ± 3.29 | 95.73 ± 0.30 | 70.85 ± 0.81 | [pth](https://github.com/YutingLi0606/Idempotent-Continual-Learning/tree/main/experiments/seq-cifar10/buffer_200/erid_seed_0.pth) | | | 500 | ER+ID | 12.02 ± 1.39 | 96.07 ± 0.19 | 75.06 ± 0.95 | [pth](https://github.com/YutingLi0606/Idempotent-Continual-Learning/tree/main/experiments/seq-cifar10/buffer_500/erid_seed_0.pth) | | CIFAR‑100 | 500 | ER+ID | 31.85 ± 3.50 | 83.45 ± 0.37 | 45.55 ± 0.66 | [pth](https://github.com/YutingLi0606/Idempotent-Continual-Learning/tree/main/experiments/seq-cifar100/buffer_500/erid_seed_0.pth) | | | 2000 | ER+ID | 18.99 ± 1.09 | 86.79 ± 0.30 | 56.15 ± 0.31 | [pth](https://github.com/YutingLi0606/Idempotent-Continual-Learning/tree/main/experiments/seq-cifar100/buffer_2000/erid_seed_0.pth) | | Tiny‑ImageNet | 4000 | ER+ID | 20.73 ± 0.72 | 74.30 ± 0.97 | 43.15 ± 1.20 | [pth](https://github.com/YutingLi0606/Idempotent-Continual-Learning/tree/main/experiments/seq-tinyimg/buffer_4000/erid_seed_0.pth) |
The checkpoints are saved under experiments folder. ## 3. Tools
mlflow visulization 1. Setup ```bash pip install mlflow ``` 2. All results are stored in mlflow under the repository mlruns. You can run mlflow ui server locally: ```bash mlflow ui ``` And then go to http://127.0.0.1:5000/#/ in your brower to see all the results from the experiments we runned and exact hyperparameters used in each run.
## 4. Citation If our project is helpful for your research, please consider citing : ``` @article{liu2026ider, title={IDER: IDempotent Experience Replay for Reliable Continual Learning}, author={Liu, Zhanwang and Li, Yuting and Gao, Haoyuan and Li, Yexin and Kong, Linghe and Sun, Lichao and Huang, Weiran}, journal={arXiv preprint arXiv:2603.00624}, year={2026} } ``` ## 5. Acknowledgement Supported by the Shanghai Municipal Special Program for Basic Research on General AI Foundation Models (Grant No. 2025SHZDZX025G03) and the SJTU Kunpeng \& Ascend Center of Excellence. This project is heavily based on [Mammoth](https://github.com/aimagelab/mammoth) and [weight-interpolation-cl](https://github.com/jedrzejkozal/weight-interpolation-cl). We sincerely appreciate the authors of the mentioned works for sharing such great library as open-source project. ✨ Feel free to contribute and reach out if you have any questions! ✨ 📧 Email: [zhanwnagliu@gmail.com](mailto:zhanwnagliu@gmail.com)