Exploring Example Influence in Continual Learning

Authors: Qing Sun, Fan Lyu, Fanhua Shang, Wei Feng, Liang Wan

NeurIPS 2022 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental Empirical results show that our algorithm significantly outperforms state-of-the-art methods on both taskand class-incremental benchmark CL datasets.
Researcher Affiliation Academia Qing Sun , Fan Lyu , Fanhua Shang, Wei Feng, Liang Wan College of Intelligence and Computing, Tianjin University {sssunqing, fanlyu, fhshang, wfeng, lwan}@tju.edu.cn
Pseudocode Yes Algorithm 1: Computation of Example Influence (Meta SP) Input: Bold, Bnew, Vold, Vnew ; // Training batches, Validation batches Output: I ; // Pareto example influence on SP ... Algorithm 2: Using Example Influence in Rehearsal-based Continual Learning. ... Algorithm 3: Training New Task
Open Source Code Yes https://github.com/SSSun Qing/Example_Influence_CL
Open Datasets Yes We use three commonly used benchmarks for evaluation: 1) Split CIFAR-10 [37]... 2) Split CIFAR-100 [37]... 3) Split Mini-Imagenet [35] is a subset of 100 classes from Image Net [9], rescaled to 32 32. Each class has 600 samples, randomly subdivided into training (80%) and test sets (20%).
Dataset Splits Yes Split Mini-Imagenet is a subset of 100 classes from Image Net [9], rescaled to 32 32. Each class has 600 samples, randomly subdivided into training (80%) and test sets (20%). We construct the SP validation sets in Meta SP by randomly sampling 10% of the seen data and 10% of the memory buffer at each training step.
Hardware Specification No No specific hardware details such as GPU/CPU models or processor types were provided in the paper text.
Software Dependencies No No specific software dependencies with version numbers (e.g., Python, PyTorch, TensorFlow versions) were mentioned in the paper.
Experiment Setup Yes We employ Res Net-18 [17] as the backbone which is trained from scratch. We use Stochastic Gradient Descent (SGD) optimizer and set the batch size 32 unchanged in order to guarantee an equal number of updates. Also, the rehearsal batch sampled from memory buffer is set to 32. We construct the SP validation sets in Meta SP by randomly sampling 10% of the seen data and 10% of the memory buffer at each training step. We set other hyper-settings following ER tricks [4], including 50 total epochs and hyper-parameters.