Sliding Windows Are Not the End: Exploring Full Ranking with Long-Context Large Language Models

• URL: http://arxiv.org/abs/2412.14574v1
• Date: Thu, 19 Dec 2024 06:44:59 GMT
• Title: Sliding Windows Are Not the End: Exploring Full Ranking with Long-Context Large Language Models
• Authors: Wenhan Liu, Xinyu Ma, Yutao Zhu, Ziliang Zhao, Shuaiqiang Wang, Dawei Yin, Zhicheng Dou,
• Abstract summary: Long-context Language Models (LLMs) enable the full ranking of all passages within a single inference. We show that full ranking with long-context LLMs can deliver superior performance in the supervised fine-tuning setting. We propose a new complete listwise label construction approach and a novel importance-aware learning objective for full ranking.
• Score: 40.21540137079309
• License:
• Abstract: Large Language Models (LLMs) have shown exciting performance in listwise passage ranking. Due to the limited input length, existing methods often adopt the sliding window strategy. Such a strategy, though effective, is inefficient as it involves repetitive and serialized processing, which usually re-evaluates relevant passages multiple times. As a result, it incurs redundant API costs, which are proportional to the number of inference tokens. The development of long-context LLMs enables the full ranking of all passages within a single inference, avoiding redundant API costs. In this paper, we conduct a comprehensive study of long-context LLMs for ranking tasks in terms of efficiency and effectiveness. Surprisingly, our experiments reveal that full ranking with long-context LLMs can deliver superior performance in the supervised fine-tuning setting with a huge efficiency improvement. Furthermore, we identify two limitations of fine-tuning the full ranking model based on existing methods: (1) sliding window strategy fails to produce a full ranking list as a training label, and (2) the language modeling loss cannot emphasize top-ranked passage IDs in the label. To alleviate these issues, we propose a new complete listwise label construction approach and a novel importance-aware learning objective for full ranking. Experiments show the superior performance of our method over baselines. Our codes are available at \url{https://github.com/8421BCD/fullrank}.

Related papers

• Self-Calibrated Listwise Reranking with Large Language Models [137.6557607279876]
  Large language models (LLMs) have been employed in reranking tasks through a sequence-to-sequence approach. This reranking paradigm requires a sliding window strategy to iteratively handle larger candidate sets. We propose a novel self-calibrated listwise reranking method, which aims to leverage LLMs to produce global relevance scores for ranking.
  arXiv  Detail & Related papers  (2024-11-07T10:31:31Z)
• RRADistill: Distilling LLMs' Passage Ranking Ability for Long-Tail Queries Document Re-Ranking on a Search Engine [2.0379810233726126]
  Large Language Models (LLMs) excel at understanding the semantic relationships between queries and documents. These queries are challenging for feedback-based rankings due to sparse user engagement and limited feedback. We propose an efficient label generation pipeline and novel sLLM training methods for both encoder and decoder models.
  arXiv  Detail & Related papers  (2024-10-08T11:28:06Z)
• FIRST: Faster Improved Listwise Reranking with Single Token Decoding [56.727761901751194]
  First, we introduce FIRST, a novel listwise LLM reranking approach leveraging the output logits of the first generated identifier to directly obtain a ranked ordering of the candidates. Empirical results demonstrate that FIRST accelerates inference by 50% while maintaining a robust ranking performance with gains across the BEIR benchmark. Our results show that LLM rerankers can provide a stronger distillation signal compared to cross-encoders, yielding substantial improvements in retriever recall after relevance feedback.
  arXiv  Detail & Related papers  (2024-06-21T21:27:50Z)
• Leveraging Passage Embeddings for Efficient Listwise Reranking with Large Language Models [17.420756201557957]
  We propose PE-Rank, leveraging the single passage embedding as a good context compression for efficient listwise passage reranking. We introduce an inference method that dynamically constrains the decoding space to these special tokens, accelerating the decoding process. Results on multiple benchmarks demonstrate that PE-Rank significantly improves efficiency in both prefilling and decoding, while maintaining competitive ranking effectiveness.
  arXiv  Detail & Related papers  (2024-06-21T03:33:51Z)
• EcoRank: Budget-Constrained Text Re-ranking Using Large Language Models [6.109188517569139]
  We study how to maximize the re-ranking performance given a budget. We propose a suite of budget-constrained methods to perform text re-ranking.
  arXiv  Detail & Related papers  (2024-02-16T18:03:42Z)
• List-aware Reranking-Truncation Joint Model for Search and Retrieval-augmented Generation [80.12531449946655]
  We propose a Reranking-Truncation joint model (GenRT) that can perform the two tasks concurrently. GenRT integrates reranking and truncation via generative paradigm based on encoder-decoder architecture. Our method achieves SOTA performance on both reranking and truncation tasks for web search and retrieval-augmented LLMs.
  arXiv  Detail & Related papers  (2024-02-05T06:52:53Z)
• OverPrompt: Enhancing ChatGPT through Efficient In-Context Learning [49.38867353135258]
  We propose OverPrompt, leveraging the in-context learning capability of LLMs to handle multiple task inputs. Our experiments show that OverPrompt can achieve cost-efficient zero-shot classification without causing significant detriment to task performance.
  arXiv  Detail & Related papers  (2023-05-24T10:08:04Z)
• Zero-Shot Listwise Document Reranking with a Large Language Model [58.64141622176841]
  We propose Listwise Reranker with a Large Language Model (LRL), which achieves strong reranking effectiveness without using any task-specific training data. Experiments on three TREC web search datasets demonstrate that LRL not only outperforms zero-shot pointwise methods when reranking first-stage retrieval results, but can also act as a final-stage reranker.
  arXiv  Detail & Related papers  (2023-05-03T14:45:34Z)

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.