Saved in:
| Main Authors: | , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2502.14100 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866916625339383808 |
|---|---|
| author | Zeng, Shenglai He, Pengfei Guo, Kai Zheng, Tianqi Lu, Hanqing Xing, Yue Liu, Hui |
| author_facet | Zeng, Shenglai He, Pengfei Guo, Kai Zheng, Tianqi Lu, Hanqing Xing, Yue Liu, Hui |
| contents | Large Language Models (LLMs) enhanced with external contexts, such as through retrieval-augmented generation (RAG), often face challenges in handling imperfect evidence. They tend to over-rely on external knowledge, making them vulnerable to misleading and unhelpful contexts. To address this, we propose the concept of context-robust LLMs, which can effectively balance internal knowledge with external context, similar to human cognitive processes. Specifically, context-robust LLMs should rely on external context only when lacking internal knowledge, identify contradictions between internal and external knowledge, and disregard unhelpful contexts. To achieve this goal, we introduce Grft, a lightweight and plug-and-play gated representation fine-tuning approach. Grft consists of two key components: a gating mechanism to detect and filter problematic inputs, and low-rank representation adapters to adjust hidden representations. By training a lightweight intervention function with only 0.0004\% of model size on fewer than 200 examples, Grft can effectively adapt LLMs towards context-robust behaviors. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2502_14100 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Towards Context-Robust LLMs: A Gated Representation Fine-tuning Approach Zeng, Shenglai He, Pengfei Guo, Kai Zheng, Tianqi Lu, Hanqing Xing, Yue Liu, Hui Computation and Language Information Retrieval Large Language Models (LLMs) enhanced with external contexts, such as through retrieval-augmented generation (RAG), often face challenges in handling imperfect evidence. They tend to over-rely on external knowledge, making them vulnerable to misleading and unhelpful contexts. To address this, we propose the concept of context-robust LLMs, which can effectively balance internal knowledge with external context, similar to human cognitive processes. Specifically, context-robust LLMs should rely on external context only when lacking internal knowledge, identify contradictions between internal and external knowledge, and disregard unhelpful contexts. To achieve this goal, we introduce Grft, a lightweight and plug-and-play gated representation fine-tuning approach. Grft consists of two key components: a gating mechanism to detect and filter problematic inputs, and low-rank representation adapters to adjust hidden representations. By training a lightweight intervention function with only 0.0004\% of model size on fewer than 200 examples, Grft can effectively adapt LLMs towards context-robust behaviors. |
| title | Towards Context-Robust LLMs: A Gated Representation Fine-tuning Approach |
| topic | Computation and Language Information Retrieval |
| url | https://arxiv.org/abs/2502.14100 |