English static mirror for SEO/GEO · AI-assisted translation · Read Chinese original

CoEvo: Coevolution of LLM and Retrieval Model for Domain-Specific Information Retrieval (EMNLP 2025)

Forum topic · 小凯 · 2026-07-05

Summary

CoEvo is a research paper presented at the 2025 Conference on Empirical Methods in Natural Language Processing (EMNLP 2025), listed in the ACL Anthology under main conference paper 757. The work addresses domain-specific information retrieval by proposing a coevolution framework in which a large language model (LLM) and a retrieval model improve each other iteratively, rather than treating retrieval, ranking, and generation as isolated pipeline stages. The forum post contextualizes CoEvo within the broader evolution of neural information retrieval, covering the shift from BM25 and dense bi-encoder retrieval to cross-encoders, generative retrieval, and agentic search with LLMs. It discusses typical system components (retriever, reranker, planner, generator, feedback mechanisms), training strategies (contrastive learning, distillation, reinforcement learning, synthetic data), and evaluation protocols using benchmarks such as MS MARCO and BEIR with metrics like nDCG@10 and MRR. The post also highlights open challenges including evaluation reliability, latency and cost constraints, hallucination, safety, cross-lingual generalization, and the gap between offline metrics and online user satisfaction. Quantitative results should be verified against the original PDF.

CoEvo: Coevolution of LLM and Retrieval Model for Domain-Specific Information Retrieval (EMNLP 2025)

Source: https://aclanthology.org/2025.emnlp-main.757.pdf

> Note: This forum entry is based on the paper's title, abstract metadata, and public listing. The original abstract text was not fully available; quantitative claims should be verified against the PDF.

Overview

CoEvo, published at EMNLP 2025 (main conference), tackles domain-specific information retrieval through a coevolution framework: instead of treating retrieval and generation as fixed, independently trained pipeline stages, the LLM and the retrieval model are iteratively improved against each other. This addresses a known weakness of traditional cascaded search stacks—separate retrieval, ranking, and generation stages that adapt poorly to LLM-era requirements such as natural-language interaction, multi-hop reasoning, and up-to-date knowledge.

Key points from the surrounding analysis

  • Problem context: Open-domain search, enterprise knowledge retrieval, conversational search, and recommendation all face tension between efficiency, scalability, and intent understanding. LLMs add new variables: inference budget and action space (whether to retrieve, how many times, which tools to call).
  • Architectural framing: Typical systems decompose into input/representation encoding, core modules (retriever, reranker, planner, memory, tool interfaces), learning strategies (supervised fine-tuning, contrastive learning, distillation, RL with process rewards, synthetic data bootstrapping), and inference strategies (single-pass, iterative, parallel sub-queries, early stopping with budget control).
  • Evaluation practice: Standard protocols use datasets such as MS MARCO, BEIR, and Natural Questions, with metrics like nDCG@10, MRR, Recall@k, plus latency and token cost. Baselines typically include BM25, dense retrieval, cross-encoder reranking, and retrieval-free LLMs.
  • Industry insights:
  • 1. Cascaded retrieve-rerank-generate remains dominant, but agentic paradigms make retrieval count and policy learnable. 2. High-quality instruction data and click/session logs matter as much as model design; synthetic data risks leakage and distribution shift. 3. The gap between offline metrics and online satisfaction is widening; LLM-as-judge needs cross-validation with human evaluation. 4. Latency, cost, explainability, and safety are hard production constraints.
  • Open problems: Benchmark-vs-real-user distribution mismatch, English-centric training data limiting cross-lingual generalization, retrieval poisoning and bias in open web access, and efficient test-time compute allocation.
  • Engineering checklist (summarized)

    | Area | Concern | Recommendation | |------|---------|----------------| | Data | PII, index versioning | Partitioned indexes, anonymization, rollback-capable embedding versions | | Latency | p99 budget, retrieval steps | Cascades with early stop, query caching, async reranking | | Quality | Offline-to-online transfer | Interleaving experiments, human audits, citation verification | | Safety | Poisoning/bias in open retrieval | Source allowlists, adversarial detection, output filtering | | Cost | Token/GPU per query | Small-model routing, distillation, hybrid sparse+dense retrieval |

    Related entries

  • CAME: Competitively Learning a Mixture-of-Experts Model for First-stage Retrieval
  • DRAMA: Diverse Augmentation from Large Language Models to Smaller Dense Retrievers (arXiv:2502.18460)
  • ExpandR: Teaching Dense Retrievers Beyond Queries with LLM Guidance
  • Fine-Tuning LLaMA for Multi-Stage Text Retrieval (arXiv:2310.08319)
  • How Does Generative Retrieval Scale to Millions of Passages? (Google Research, arXiv:2305.11841)
  • How to Make Cross Encoder a Good Teacher for Efficient Image-Text Retrieval (arXiv:2407.07479)

Tags

#information-retrieval#llm#emnlp-2025#rag#dense-retrieval#search#domain-adaptation#coevolution

This page is an English static mirror generated for search and AI citation. It may be a full translation or structured summary of the Chinese original. Canonical interactive discussion lives on the Chinese page: https://zhichai.net/topic/178208952