arXiv Machine Learning June 16, 2026 By Gustav Olaf Yunus Laitinen-Fredriksson Lundstrom-Imanov

Mechanistic Analysis of Catastrophic Forgetting in Large Language Models During Continual Fine-tuning

Read the original on arXiv Machine Learning →

arXiv:2601. 18699v2 Announce Type: replace Abstract: Sequential fine-tuning of Large Language Models (LLMs) adaptation to target tasks often triggers catastrophic forgetting, where the acquisition of novel target skills degrades ancestral capabilities.

Summary generated by The Flow from the publisher's feed. The full article lives at arXiv Machine Learning.

arXiv AI

Jul 20

PASs-MoE: Mitigating Misaligned Co-drift among Router and Experts via Pathway Activation Subspaces for Continual Learning

arXiv:2601. 13020v2 Announce Type: replace-cross Abstract: Continual instruction tuning (CIT) requires multimodal large language models (MLLMs) to adapt to a stream of tasks without forgetting prior capabilities.

By Zhiyan Hou, Haiyun Guo, Haokai Ma, Yandu Sun, Yonghui Yang, Jinqiao Wang

llms fine-tuning multimodal benchmarks

arXiv Machine Learning

Jul 30

The Art of Not Forgetting A Local Learning Architecture for Continual Learning

arXiv:2607. 26523v1 Announce Type: new Abstract: We introduce CMP (Cognitive Memory Primitive), a continual-learning architecture that repre?

By Ashmith Atmuri, Yashaswini Rao Bhogarajula

llms benchmarks

arXiv Machine Learning

6d ago

Drift and Dependence: Layer-wise Information-Theoretic Bounds for Replay-Based Continual Learning

arXiv:2608. 11690v1 Announce Type: new Abstract: Continual learning must absorb new tasks without erasing old ones, and replay---mixing a small buffer of past examples into current training---is among the most effective remedies for catastrophic forgetting.

By Tieliang Gong, Zhongbo Zhang, Wen Wen, Yong-Jin Liu

benchmarks safety

arXiv AI

Jun 10

Routing-Aware Expert Calibration for Machine Unlearning in Mixture-of-Experts Language Models

arXiv:2606. 10338v1 Announce Type: cross Abstract: Machine unlearning is increasingly important for large language models, yet unlearning in Mixture-of-Experts (MoE) architectures remains underexplored.

By Jingyi Xie, Yijun Lin, Yinjiang Xiong, Zhikun Zhang, Sai Li

llms

arXiv AI

Jun 18

From Memorization to Parameter Interference: How Overtraining Experts Harms Model Merging

arXiv:2506. 14126v2 Announce Type: replace-cross Abstract: Modern deep learning is increasingly characterized by the use of open-weight foundation models that can be fine-tuned on specialized datasets.

By Stefan Horoi, Guy Wolf, Eugene Belilovsky, Gintare Karolina Dziugaite

fine-tuning

arXiv Machine Learning

Jul 15

Inference-Time Machine Unlearning via Gated Activation Redirection

arXiv:2605. 12765v3 Announce Type: replace Abstract: Large Language Models memorize vast amounts of training data, raising concerns regarding privacy, copyright infringement, and safety.

By Vin\'icius Conte Turani, Ot\'avio Parraga, Jo\~ao Vitor Boer Abitante, Kristen K. Arguello, Joana Pasquali, Ramiro N. Barros, Flavio du Pin Calmon, Christian Mattjie, Rodrigo C. Barros, Lucas S. Kupssinsk\"u

llms efficiency safety

Related stories

PASs-MoE: Mitigating Misaligned Co-drift among Router and Experts via Pathway Activation Subspaces for Continual Learning

The Art of Not Forgetting A Local Learning Architecture for Continual Learning

Drift and Dependence: Layer-wise Information-Theoretic Bounds for Replay-Based Continual Learning

Routing-Aware Expert Calibration for Machine Unlearning in Mixture-of-Experts Language Models

From Memorization to Parameter Interference: How Overtraining Experts Harms Model Merging

Inference-Time Machine Unlearning via Gated Activation Redirection