← Back to all news
arXiv Machine Learning By Yu Sun, Junwei Zhou, Zuodong Xiang, Yike Zhang, Pengcheng Xu, Xinyu Guan, Ruoyun Ma, Hailu Xu

Static Gradient Attribution Underperforms a Density-Matched Random Mask on Loss-Based Forgetting Within LoRA B-Matrix

Read the original on arXiv Machine Learning →

The study examines parameter‑efficient continual learning with LoRA, comparing gradient‑based attribution masks to random masks for selecting adapter parameters. It finds that when both methods are restricted to the same parameter pool, uniform random masks consistently outperform gradient attribution in reducing loss‑based backward transfer across multiple seeds on the Qwen2.5‑3B model, especially on the medical task. This suggests that random masking may better balance stability and plasticity in this setting.

Machine-generated by The Flow from the publisher's headline and feed description — not written or checked by a human. The full article lives at arXiv Machine Learning.

Related stories

arXiv AI
Jul 22

Soft-TransFormers for Continual Learning

arXiv:2411. 16073v4 Announce Type: replace-cross Abstract: Inspired by the Well-initialized Lottery Ticket Hypothesis (WLTH), we introduce Soft-TransFormers (Soft-TF), a continual learning framework that adapts a frozen pre-trained Transformer through task-specific soft subnetworks: real-valued multiplicative masks over the query, key, value, and output projections of selected self-attention layers.

By Haeyong Kang, Chang D. Yoo
llmsfine-tuningbenchmarks
More like this →
arXiv AI
Jun 3

PURGE: Projected Unlearning via Retain-Guided Erasure

arXiv:2606. 03808v1 Announce Type: cross Abstract: We propose PURGE, a machine unlearning algorithm built on a simple but an under-exploited observation: continual learning (CL) and machine unlearning (MU) which are fundamentally dual problems.

By Vedant Jawandhia, Daksh Ahuja, Ghufran Alam Siddiqui, Prashant Trivedi, Yash Sinha, Pratik Narang
safety
More like this →
arXiv Machine Learning
Oct 5

Planning to Learn

The paper introduces a new loss function called the horizon loss for training classifiers. It argues that the exact policy gradient used in reinforcement learning is myopic, whereas cross‑entropy is patient, and the horizon loss interpolates between the two by truncating the total error at the remaining learning. Experiments on MNIST and ImageNet with ResNet and ViT models show that horizon loss consistently improves top‑1 accuracy over cross‑entropy, especially when label noise is present.

By Ian Osband
llmsreinforcement-learning
More like this →
arXiv Machine Learning
5d ago

On the Origin and Evolution of Training-Inference Mismatch in LLM Reinforcement Learning

The paper investigates the training‑inference mismatch that arises when large language models are trained with on‑policy reinforcement learning but inference is performed by a separate engine. It identifies two sources of mismatch—numerical discrepancies in the engine and policy lag due to stale weights—and shows that correcting one source can worsen the other. The authors propose a discrepancy‑lag correction (DLC) that centers numerical errors, clips lag ratios, and resynchronizes the sampler, achieving stable training on Qwen models.

By Yu Li, Yufa Zhou, Tian Lan
llmsreinforcement-learningsafety
More like this →