Learning from Synthetic Data Improves Multi-hop Reasoning

• URL: http://arxiv.org/abs/2603.02091v1
• Date: Mon, 02 Mar 2026 17:08:43 GMT
• Title: Learning from Synthetic Data Improves Multi-hop Reasoning
• Abstract summary: Reinforcement Learning (RL) has been shown to significantly boost reasoning capabilities of large language models (LLMs) in math, coding, and multi-hop reasoning tasks.<n>In this work, we investigate a cheaper alternative: RL fine-tuning on rule-generated synthetic data for multi-hop reasoning tasks.
• Score: 31.79471776662371
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Reinforcement Learning (RL) has been shown to significantly boost reasoning capabilities of large language models (LLMs) in math, coding, and multi-hop reasoning tasks. However, RL fine-tuning requires abundant high-quality verifiable data, often sourced from human annotations, generated from frontier LLMs, or scored by LLM-based verifiers. All three have considerable limitations: human-annotated datasets are small and expensive to curate, LLM-generated data is hallucination-prone and costly, and LLM-based verifiers are inaccurate and slow. In this work, we investigate a cheaper alternative: RL fine-tuning on rule-generated synthetic data for multi-hop reasoning tasks. We discover that LLMs fine-tuned on synthetic data perform significantly better on popular real-world question-answering benchmarks, despite the synthetic data containing only fictional knowledge. On stratifying performance by question difficulty, we find that synthetic data teaches LLMs to compose knowledge -- a fundamental and generalizable reasoning skill. Our work highlights rule-generated synthetic reasoning data as a free and scalable resource to improve LLM reasoning capabilities.

Related papers

• Verifying the Verifiers: Unveiling Pitfalls and Potentials in Fact Verifiers [59.168391398830515]
  We evaluate 12 pre-trained LLMs and one specialized fact-verifier, using a collection of examples from 14 fact-checking benchmarks.<n>We highlight the importance of addressing annotation errors and ambiguity in datasets.<n> frontier LLMs with few-shot in-context examples, often overlooked in previous works, achieve top-tier performance.
  arXiv  Detail & Related papers  (2025-06-16T10:32:10Z)
• SciDA: Scientific Dynamic Assessor of LLMs [18.120612887031598]
  Large Language Models (LLMs) reasoning capabilities enable them to solve scientific problems with enhanced efficacy.<n>Existing benchmarks either confront the risk of data contamination or lack involved disciplines.<n>We propose SciDA, a multidisciplinary benchmark that consists exclusively of over 1k Olympic-level numerical computation problems.
  arXiv  Detail & Related papers  (2025-06-15T16:57:14Z)
• LightPROF: A Lightweight Reasoning Framework for Large Language Model on Knowledge Graph [57.382255728234064]
  Large Language Models (LLMs) have impressive capabilities in text understanding and zero-shot reasoning.<n> Knowledge Graphs (KGs) provide rich and reliable contextual information for the reasoning process of LLMs.<n>We propose a novel Lightweight and efficient Prompt learning-ReasOning Framework for KGQA (LightPROF)
  arXiv  Detail & Related papers  (2025-04-04T03:03:47Z)
• Search-R1: Training LLMs to Reason and Leverage Search Engines with Reinforcement Learning [50.419872452397684]
  Search-R1 is an extension of reinforcement learning for reasoning frameworks.<n>It generates search queries during step-by-step reasoning with real-time retrieval.<n>It improves performance by 41% (Qwen2.5-7B) and 20% (Qwen2.5-3B) over various RAG baselines.
  arXiv  Detail & Related papers  (2025-03-12T16:26:39Z)
• Agent Trading Arena: A Study on Numerical Understanding in LLM-Based Agents [69.58565132975504]
  Large language models (LLMs) have demonstrated remarkable capabilities in natural language tasks.<n>We present the Agent Trading Arena, a virtual zero-sum stock market in which LLM-based agents engage in competitive multi-agent trading.
  arXiv  Detail & Related papers  (2025-02-25T08:41:01Z)
• Escaping Collapse: The Strength of Weak Data for Large Language Model Training [15.77316232527746]
  We develop a theoretical framework to investigate how much curation is needed in order to ensure that LLM performance continually improves.<n>We describe a training procedure that converges to an optimal LLM even if almost all of the non-synthetic training data is of poor quality.
  arXiv  Detail & Related papers  (2025-02-13T03:20:37Z)
• FactCG: Enhancing Fact Checkers with Graph-Based Multi-Hop Data [13.108807408880645]
  We propose a novel approach for synthetic data generation, CG2C, that leverages multi-hop reasoning on context graphs extracted from documents.<n>Our fact checker model, FactCG, demonstrates improved performance with more connected reasoning, using the same backbone models.
  arXiv  Detail & Related papers  (2025-01-28T18:45:07Z)
• SELF-GUIDE: Better Task-Specific Instruction Following via Self-Synthetic Finetuning [70.21358720599821]
  Large language models (LLMs) hold the promise of solving diverse tasks when provided with appropriate natural language prompts. We propose SELF-GUIDE, a multi-stage mechanism in which we synthesize task-specific input-output pairs from the student LLM. We report an absolute improvement of approximately 15% for classification tasks and 18% for generation tasks in the benchmark's metrics.
  arXiv  Detail & Related papers  (2024-07-16T04:41:58Z)
• DARG: Dynamic Evaluation of Large Language Models via Adaptive Reasoning Graph [70.79413606968814]
  We introduce Dynamic Evaluation of LLMs via Adaptive Reasoning Graph Evolvement (DARG) to dynamically extend current benchmarks with controlled complexity and diversity. Specifically, we first extract the reasoning graphs of data points in current benchmarks and then perturb the reasoning graphs to generate novel testing data. Such newly generated test samples can have different levels of complexity while maintaining linguistic diversity similar to the original benchmarks.
  arXiv  Detail & Related papers  (2024-06-25T04:27:53Z)
• Q*: Improving Multi-step Reasoning for LLMs with Deliberative Planning [53.6472920229013]
  Large Language Models (LLMs) have demonstrated impressive capability in many natural language tasks. LLMs are prone to produce errors, hallucinations and inconsistent statements when performing multi-step reasoning. We introduce Q*, a framework for guiding LLMs decoding process with deliberative planning.
  arXiv  Detail & Related papers  (2024-06-20T13:08:09Z)
• MiniCheck: Efficient Fact-Checking of LLMs on Grounding Documents [62.02920842630234]
  We show how to build small fact-checking models that have GPT-4-level performance but for 400x lower cost. We do this by constructing synthetic training data with GPT-4, which involves creating realistic yet challenging instances of factual errors. For evaluation, we unify datasets from recent work on fact-checking and grounding LLM generations into a new benchmark, LLM-AggreFact.
  arXiv  Detail & Related papers  (2024-04-16T17:59:10Z)
• Under the Surface: Tracking the Artifactuality of LLM-Generated Data [21.002983022237604]
  This work delves into the expanding role of large language models (LLMs) in generating artificial data. To the best of our knowledge, this is the first study to aggregate various types of LLM-generated text data. Despite artificial data's capability to match human performance, this paper reveals significant hidden disparities.
  arXiv  Detail & Related papers  (2024-01-26T07:53:27Z)
• Causal Reasoning and Large Language Models: Opening a New Frontier for Causality [29.433401785920065]
  Large language models (LLMs) can generate causal arguments with high probability. LLMs may be used by human domain experts to save effort in setting up a causal analysis.
  arXiv  Detail & Related papers  (2023-04-28T19:00:43Z)

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.