用高斯混合模型优化记忆重放，提升大模型持续学习时的记性。
Enhancing Memory Recall in LLMs with Gauss-Tin: A Hybrid Instructional and Gaussian Replay Approach
- 结合高斯混合模型与指令引导，智能选择重放样本。
- 相比传统方法，记忆保留率提升6%。
- 适合需要长期学习新知识又不丢旧知识的场景。
尽管大型语言模型（LLMs）取得显著进展，但灾难性遗忘仍是主要挑战，即模型在学习新知识时会丢失已有知识。持续学习（CL）策略成为潜在解决方案，其中基于重放的方法在保留知识方面表现更优。本文提出Gauss-Tin，一种将重放策略与高斯混合模型结合的新方法，通过高质量样本选择增强训练效果，并辅以指令引导促进过往学习内容的生成。该方法旨在通过有策略地强化重要旧知识，同时容纳新信息，从而提升大模型的记忆保持能力。实验结果表明，相较于传统方法，记忆保留指标提升6%，验证了Gauss-Tin在缓解大模型灾难性遗忘方面的有效性。本研究凸显了混合模型在动态学习环境中增强大模型鲁棒性与适应性的潜力。
原文摘要 · Abstract (English)
Despite the significant advancements in Large Language Models (LLMs), catastrophic forgetting remains a substantial challenge, where models lose previously acquired knowledge upon learning new information. Continual learning (CL) strategies have emerged as a potential solution to this problem, with replay-based techniques demonstrating superior performance in preserving learned knowledge. In this context, we introduce Gauss-Tin, a novel approach that integrates the replay strategy with a Gaussian mixture model to enhance the quality of sample selection during training, supplemented by instructional guidance to facilitate the generation of past learning. This method aims to improve LLMs' retention capabilities by strategically reinforcing important past learnings while accommodating new information. Our experimental results indicate a promising 6\% improvement in retention metrics over traditional methods, suggesting that Gauss-Tin is an effective strategy for mitigating catastrophic forgetting in LLMs. This study underscores the potential of hybrid models in enhancing the robustness and adaptability of LLMs in dynamic learning environments.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。