---
title: Multi-hop Dense Retrieval
url: https://www.emergentmind.com/topics/multi-hop-dense-retrieval-mdr
type: topic
---

# Multi-hop Dense Retrieval

Multi-hop Dense Retrieval (MDR) is a retrieval paradigm for answering complex information-seeking queries requiring reasoning over multiple discrete passages in a large corpus. MDR extends standard dense retrieval architectures by enabling iterative retrieval: at each hop, the system dynamically refines the query by integrating newly acquired evidence, thereby constructing a chain—or latent graph—of documents that collectively provide the necessary information to resolve the initial query. Research on MDR centers on neural methods—primarily dual-encoder models—optimized through multi-hop adaptations of contrastive learning, as well as decompositional and end-to-end strategies that address both efficiency and generalization challenges in open-domain multi-hop question answering (QA) and complex reasoning tasks.

## 1. Formal Definition and Core Principles

MDR models the evidence retrieval process as a sequence of dependent retrieval steps. The retrieval model seeks a chain of documents $[d_1, d_2, ..., d_L]$ such that each $d_t$ is selected from a corpus $\mathcal{D}$ conditioned on the question $q$ and the evidence $d_1, ..., d_{t-1}$ collected in previous hops. At each hop, the retriever forms a context-augmented query $q_t$ by composition—typically by concatenation or embedding-based fusion—of the question and previously retrieved evidence:
\[
q_t = \text{Compose}(q, d_1, ..., d_{t-1})
\]
Retrieval operates as maximum inner product search (MIPS) in a dense embedding space. Formally, at each hop,
\[
d_t = \arg\max_{d \in \mathcal{D}} \langle E_Q(q_t), E_D(d) \rangle
\]
where $E_Q$ and $E_D$ are neural encoders for queries and documents, and $\langle \cdot, \cdot \rangle$ denotes inner product or, in some variants, cosine similarity [2009.12756, 2106.08433, 2106.09983].

Score aggregation across hops is typically additive or multiplicative, and candidate reasoning chains are expanded in a beam search or scored jointly over passage sequences. The ability to recursively update the retrieval query as a function of prior evidence is fundamental to MDR’s effectiveness in scenarios where no single passage directly answers the question.

## 2. Methodological Taxonomy and Modeling Choices

### 2.1 Dual-Encoder Architectures

Most MDR systems employ dual-encoder (bi-encoder) architectures for efficient large-scale retrieval [2009.12756, 2106.08433, 2104.05883, 2106.09983]. For each text input $x$, the encoder produces a dense vector $v_x$ (e.g., via the [CLS] token of a Transformer, projected to dimension $d$). Training follows hard negative contrastive learning:
\[
\mathcal{L}_{\text{CL}} = -\log \frac{\exp(\operatorname{sim}(v_{q_t}, v_{d_t^+})/\tau)}{\sum_j\exp(\operatorname{sim}(v_{q_t}, v_{d_j})/\tau)}
\]
with in-batch and/or mined negatives.

### 2.2 Query Update Mechanisms

Critical to MDR is query composition:
- **Simple concatenation**: $q_{t} = [q; d_1; ...; d_{t-1}]$ (input-level) [2009.12756, 2106.09983].
- **Embedding fusion**: $z_{q}^{(t)} = \text{MLP}([z_{q}^{(t-1)}; z_d])$ or $z_q^{(t)} = z_q^{(t-1)} + \alpha z_d$ [2104.05883].
- **Attention-based context fusion**: $\mathbf{D}_{\text{agg}} = \sum_{i} \alpha_i d_i$ with $\alpha_i$ via softmax over similarity [2506.16037].

### 2.3 Retrieval Path Search and Beam Search

Beam search expands multiple candidate reasoning chains, allowing the model to recover from early retrieval errors and explore alternative multi-hop paths. Each beam element holds a query state, accumulated score, and path:
``` 
For t = 1 to H:
  For each (path, q, score) in beam:
    Retrieve top-M passages; expand; prune to top-K beams [2104.05883]
```
This contrasts with greedy iterative retrieval, which follows a single chain [2009.12756].

### 2.4 Decomposition-Free versus Decomposition-Based

Decomposition-free MDR directly builds multi-hop evidence chains by recursive query updating and retrieval.
Decomposition-based approaches, by contrast, rely on rule-based or generatively-produced sub-questions and retrieve each hop conditioned on sub-question outputs, often via external LLMs or complex planner–executor structures [2508.16438].

## 3. Supervision, Optimization, and Weak Labeling

### 3.1 Contrastive and Mixed Objectives

While InfoNCE-style contrastive loss is standard, several variants integrate additional objectives:
- **Mixed classification loss**: e.g., three-way label prediction on (claim, evidence) pairs (SUPPORTS, REFUTES, NEI) [2403.14074].
- **End-to-end reinforcement learning via policy optimization**: OPERA’s MAPGRPO algorithm trains planning and retrieval agents sequentially with group-relative rewards [2508.16438].

### 3.2 Weak Supervision and Synthetic Data

Human annotation for multi-hop retrieval is exceptionally costly due to the combinatorial space of hop sequences. LOUVRE mitigates this via large-scale pseudo-labels generated by bridge-entity rephrasing on Wikipedia hyperlinks, facilitating massive weakly-supervised pretraining [2106.09983].

### 3.3 Posterior Regularization and Distillation

MoPo regularizes the prior (inference) retriever toward a posterior (oracle) model at each hop using a KL divergence term, updating the posterior model by an exponential moving average of the prior parameters to stabilize distillation [2502.20399]. This addresses the instability and performance gap when transferring strong posterior information (e.g., answer-aware queries) to real-world inference.

### 3.4 Label-Free Retriever Training with LLM Signals

ReSCORE dispenses with explicit hop-level labels, instead extracting per-passage pseudo-labels from LLM token probabilities that jointly reflect both relevance (p(q|d)) and answer consistency (p(a|q,d)), and updating the retriever to optimize KL(Q_LLM‖P_θ) in an iterative RAG loop [2505.21250]. This increases adaptability in domains where annotations are scarce.

## 4. Architectural and Efficiency Advances

### 4.1 Condensed and Focused Retrieval

Baleen addresses exponential context growth by condensing each hop’s retrieved evidence into a compact set of relevant sentences, using a two-stage condenser coupled with a focused late-interaction retriever (FLIPR) for efficient multi-hop expansion without loss of retrieval accuracy [2101.00436]. FLIPR applies token-level max-similarity between query and passage representations, focusing on subsets to better match disjoint multi-hop needs.

### 4.2 Decomposition-Free Generative Models

GRITHopper integrates dense retrieval training and causal language modeling on a unified decoder architecture, employing “retrieve–evaluate–act” instructions, post-retrieval final answer conditioning, and reward modeling. Unlike pure bi-encoders, this approach supports end-to-end differentiation and OOD generalization to longer reasoning chains [2503.07519].

### 4.3 Planner–Executor and Reinforcement Learning

OPERA explicitly models question decomposition, retrieval, and rewriting with specialized agent modules—Goal Planning, Analysis-Answer, Rewrite—trained via MAPGRPO. This orchestrates sub-goal formation and execution with dense retrieval tightly coupled to fine-grained reasoning rewards, allowing robust multi-hop search even for out-of-template queries [2508.16438].

## 5. Empirical Results and Benchmarking

MDR establishes state-of-the-art retrieval and QA performance across multiple open-domain and fact verification benchmarks:
- On HotpotQA, 2WikiMultiHopQA, FEVER, MuSiQue, and related datasets, MDR systems consistently outperform prior sparse, entity-linking, and graph-based methods [2009.12756, 2104.05883, 2403.14074, 2503.07519, 2506.16037].
- In HotpotQA 2-hop settings, classical MDR achieves R@2 ≈ 66, R@20 ≈ 80 (recall of both gold supports), with enhanced recall and end-to-end joint F1 using hybrid, beam, or condensed retrieval methods [2009.12756, 2106.08433, 2104.05883, 2101.00436].
- MR.COD demonstrates MDR’s impact beyond QA, substantially boosting recall and relation extraction F1 for cross-document RE (e.g., +7.5 absolute F1 over snippets baselines) [2212.10786].
- For strong generalization, decomposition-free models (GRITHopper) and planner–executor frameworks (OPERA) show smaller performance drops and high relative My@k on out-of-domain 4-hop QA tasks [2503.07519, 2508.16438].

Empirical ablations reveal beam size, negative mining, and query composition are critical. Weakly-supervised or LLM-pseudo labeling enables MDR to scale efficiently without human hop-level annotation [2106.09983, 2505.21250].

## 6. Scalability, Efficiency, and Limitations

The principal efficiency innovation of MDR is the use of dual-encoder architectures with offline passage indexing and sub-linear ANN search (e.g., FAISS HNSW), resulting in inference times orders of magnitude lower than rerank- or cross-attention-based pipelines [2009.12756, 2106.08433, 2101.00436]. Condensation techniques and compact context fusion further mitigate the context explosion as the number of hops increases [2101.00436, 2503.07519].

However, MDR’s susceptibility to error propagation (early-hop misretrieval), limited ability to self-correct, and challenges in scaling query composition to $T\gg2$ hops remain active research areas. Non-differentiable or multi-stage decompositional strategies can induce computation overhead and break end-to-end learning [2503.07519, 2508.16438]. Integration of sparse (BM25) and dense signals remains a promising direction for robust and resource-constrained deployment [2106.08433].

## 7. Future Directions and Open Challenges

- **Dynamic hop counting, stopping, and backtracking**: Rather than fixing the number of hops, future MDR should dynamically determine when to stop or revisit earlier steps [2503.07519].
- **Hybrid lexical–dense architectures**: New work targets seamless fusion of lexical matching and dense semantic retrieval in a single network [2106.08433, 2403.14074].
- **Multi-task and mixed-objective learning**: Incorporating verification classification, relevance, and contrastive losses in joint training yields further gains [2403.14074].
- **Reinforcement and instruction-based approaches**: Architectures that couple high-level planning, reasoning, and retrieval via policy optimization or generative instructions are gaining traction [2508.16438]. 
- **Label-free and weak supervision**: Exploiting LLMs for pseudo-labels or synthetic data generation expands MDR applicability to domains with scarce annotation [2106.09983, 2505.21250].
- **Plug-and-play compositional retrievers**: Robustness to distributional shift and higher-hop reasoning demands decomposition-free models with strong generalization [2503.07519].

MDR is a rapidly developing area at the intersection of representation learning, open-domain retrieval, and complex multi-step reasoning, with a rich set of methods and empirical validation spanning question answering, fact verification, and cross-document relation extraction.

Source: https://www.emergentmind.com/topics/multi-hop-dense-retrieval-mdr