Utilize the Flow before Stepping into the Same River Twice: Certainty Represented Knowledge Flow for Refusal-Aware Instruction Tuning

• URL: http://arxiv.org/abs/2410.06913v2
• Date: Mon, 18 Nov 2024 13:15:41 GMT
• Title: Utilize the Flow before Stepping into the Same River Twice: Certainty Represented Knowledge Flow for Refusal-Aware Instruction Tuning
• Authors: Runchuan Zhu, Zhipeng Ma, Jiang Wu, Junyuan Gao, Jiaqi Wang, Dahua Lin, Conghui He,
• Abstract summary: Refusal-Aware Instruction Tuning (RAIT) enables Large Language Models (LLMs) to refuse to answer unknown questions. RAIT modifies training samples based on the correctness of the initial LLM's response. This crude approach can cause LLMs to excessively refuse answering questions they could have correctly answered.
• Score: 68.57166425493283
• License:
• Abstract: Refusal-Aware Instruction Tuning (RAIT) enables Large Language Models (LLMs) to refuse to answer unknown questions. By modifying responses of unknown questions in the training data to refusal responses such as "I don't know", RAIT enhances the reliability of LLMs and reduces their hallucination. Generally, RAIT modifies training samples based on the correctness of the initial LLM's response. However, this crude approach can cause LLMs to excessively refuse answering questions they could have correctly answered, the problem we call over-refusal. In this paper, we explore two primary causes of over-refusal: Static conflict occurs when similar samples within the LLM's feature space receive differing supervision signals (original vs. modified "I don't know"). Dynamic conflict, on the other hand, emerges as the LLM's knowledge evolves during SFT, allowing it to answer questions that were previously unanswerable. Yet, these now-answerable training samples still retain the original "I don't know" supervision signals based on the initial LLM state, resulting in inconsistencies. These conflicts cause the trained LLM to misclassify known questions as unknown, resulting in over-refusal. To address this issue, we introduce Certainty Represented Knowledge Flow for Refusal-Aware Instructions Tuning (CRaFT). CRaFT centers on two main contributions: First, we additionally incorporate response certainty to selectively filter and modify data, reducing static conflicts. Second, we implement preliminary rehearsal training to characterize changes in the LLM's knowledge state, which helps mitigate dynamic conflicts during the fine-tuning process. We conducted extensive experiments on open-ended question answering and multiple-choice question task. Experiment results show that CRaFT can improve LLM's overall performance during the RAIT process. Source code and training data will be released at Github.

Related papers

• How Much Knowledge Can You Pack into a LoRA Adapter without Harming LLM? [55.33467849079774]
  Low-rank adaptation (LoRA) is a popular and efficient training technique for updating or domain-specific adaptation of Large Language Models. We investigate how new facts can be incorporated into the LLM using LoRA without compromising the previously learned knowledge.
  arXiv  Detail & Related papers  (2025-02-20T12:31:03Z)
• Post-training an LLM for RAG? Train on Self-Generated Demonstrations [18.8726880078299]
  Large language models (LLMs) often struggle with knowledge intensive NLP tasks. Retrieval augmented generation (RAG) allows the model to leverage in-context information. We propose a recipe for training RAG-enabled LLMs using self-generated demonstrations.
  arXiv  Detail & Related papers  (2025-02-14T23:00:49Z)
• Invar-RAG: Invariant LLM-aligned Retrieval for Better Generation [43.630437906898635]
  We propose a novel two-stage fine-tuning architecture called Invar-RAG. In the retrieval stage, an LLM-based retriever is constructed by integrating LoRA-based representation learning. In the generation stage, a refined fine-tuning method is employed to improve LLM accuracy in generating answers based on retrieved information.
  arXiv  Detail & Related papers  (2024-11-11T14:25:37Z)
• Are LLMs Aware that Some Questions are not Open-ended? [58.93124686141781]
  We study whether Large Language Models are aware that some questions have limited answers and need to respond more deterministically. The lack of question awareness in LLMs leads to two phenomena: (1) too casual to answer non-open-ended questions or (2) too boring to answer open-ended questions.
  arXiv  Detail & Related papers  (2024-10-01T06:07:00Z)
• From Yes-Men to Truth-Tellers: Addressing Sycophancy in Large Language Models with Pinpoint Tuning [91.79567270986901]
  Large Language Models (LLMs) tend to prioritize adherence to user prompts over providing veracious responses. Recent works propose to employ supervised fine-tuning (SFT) to mitigate the sycophancy issue. We propose a novel supervised pinpoint tuning (SPT), where the region-of-interest modules are tuned for a given objective.
  arXiv  Detail & Related papers  (2024-09-03T07:01:37Z)
• From Distributional to Overton Pluralism: Investigating Large Language Model Alignment [82.99849359892112]
  We re-examine previously reported reductions in response diversity post-alignment. Our analysis suggests that an apparent drop in the diversity of responses is largely explained by quality control and information aggregation. Findings indicate that current alignment techniques capture but do not extend the useful subset of assistant-like base LLM behavior.
  arXiv  Detail & Related papers  (2024-06-25T16:32:33Z)
• Rephrase and Respond: Let Large Language Models Ask Better Questions for Themselves [57.974103113675795]
  We present a method named Rephrase and Respond' (RaR) which allows Large Language Models to rephrase and expand questions posed by humans. RaR serves as a simple yet effective prompting method for improving performance. We show that RaR is complementary to the popular Chain-of-Thought (CoT) methods, both theoretically and empirically.
  arXiv  Detail & Related papers  (2023-11-07T18:43:34Z)
• Knowing What LLMs DO NOT Know: A Simple Yet Effective Self-Detection Method [36.24876571343749]
  Large Language Models (LLMs) have shown great potential in Natural Language Processing (NLP) tasks. Recent literature reveals that LLMs generate nonfactual responses intermittently. We propose a novel self-detection method to detect which questions that a LLM does not know that are prone to generate nonfactual results.
  arXiv  Detail & Related papers  (2023-10-27T06:22:14Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.