FreeMOCA enables memory-free continual learning for malicious code analysis via adaptive layer-wise interpolation between warm-started task optima, outperforming baselines on EMBER and AZ benchmarks with up to 42% accuracy gains.
Analyzing and reducing catastrophic forgetting in parameter efficient tuning.arXiv preprint arXiv:2402.18865
5 Pith papers cite this work. Polarity classification is still indexing.
5
Pith papers citing it
citation-role summary
background 3
citation-polarity summary
roles
background 3polarities
background 3representative citing papers
Self-evolving LLM agents exhibit capability erosion under continual adaptation, which Capability-Preserving Evolution mitigates by raising retained simple-task performance from 41.8% to 52.8% in workflow evolution under GPT-5.1.
Empirical study on five LLMs finds pretrained-to-aligned paths yield bigger gains over baseline than finetuned-to-aligned paths, though absolute accuracy remains lower for pretrained starts.
The paper claims a selective fine-tuning method that identifies and freezes core parameters to mitigate catastrophic forgetting in LLMs while improving domain adaptation, shown in experiments with GPT-J and LLaMA-3.
citing papers explorer
-
FreeMOCA: Memory-Free Continual Learning for Malicious Code Analysis
FreeMOCA enables memory-free continual learning for malicious code analysis via adaptive layer-wise interpolation between warm-started task optima, outperforming baselines on EMBER and AZ benchmarks with up to 42% accuracy gains.
-
Do Self-Evolving Agents Forget? Capability Degradation and Preservation in Lifelong LLM Agent Adaptation
Self-evolving LLM agents exhibit capability erosion under continual adaptation, which Capability-Preserving Evolution mitigates by raising retained simple-task performance from 41.8% to 52.8% in workflow evolution under GPT-5.1.
-
Reward-Free Code Alignment from Pretrained or Fine-Tuned LLM: Unpacking the Trade-offs for Code Generation
Empirical study on five LLMs finds pretrained-to-aligned paths yield bigger gains over baseline than finetuned-to-aligned paths, though absolute accuracy remains lower for pretrained starts.
-
Efficient Task Adaptation in Large Language Models via Selective Parameter Optimization
The paper claims a selective fine-tuning method that identifies and freezes core parameters to mitigate catastrophic forgetting in LLMs while improving domain adaptation, shown in experiments with GPT-J and LLaMA-3.