Search:
Match:
2 results
Paper#LLM🔬 ResearchAnalyzed: Jan 3, 2026 16:28

AFA-LoRA: Enhancing LoRA with Non-Linear Adaptations

Published:Dec 27, 2025 04:12
1 min read
ArXiv

Analysis

This paper addresses a key limitation of LoRA, a popular parameter-efficient fine-tuning method: its linear adaptation process. By introducing AFA-LoRA, the authors propose a method to incorporate non-linear expressivity, potentially improving performance and closing the gap with full-parameter fine-tuning. The use of an annealed activation function is a novel approach to achieve this while maintaining LoRA's mergeability.
Reference

AFA-LoRA reduces the performance gap between LoRA and full-parameter training.

Research#LLM🔬 ResearchAnalyzed: Jan 10, 2026 07:49

RevFFN: Efficient Fine-Tuning of Mixture-of-Experts LLMs with Reversible Blocks

Published:Dec 24, 2025 03:56
1 min read
ArXiv

Analysis

The research on RevFFN presents a promising approach to reduce memory consumption during the fine-tuning of large language models. The use of reversible blocks to achieve memory efficiency is a significant contribution to the field of LLM training.
Reference

The paper focuses on memory-efficient full-parameter fine-tuning of Mixture-of-Experts (MoE) LLMs with Reversible Blocks.