# Understanding and Improving Zero-shot Multi-hop Reasoning in Generative Question Answering

arXiv: https://arxiv.org/abs/2210.04234 · HTML: https://mlsanity.com/paper/2210.04234/

## Key facts

- Cited by 7 papers in the map (data through 14 Oct 2026).
- Placed in Knowledge- and tool-augmented reasoning, one of 17 directions in LLM reasoning.
- Builds on 36 of its 42 references that are in the map.
- First posted on arXiv 9 Oct 2022.

## Placed in

- LLM reasoning / [Knowledge- and tool-augmented reasoning](https://mlsanity.com/lr/knowledge-and-tool-augmented-reasoning/)

## Builds on (most cited first, 30 of 36)

- [BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding](https://mlsanity.com/paper/1810.04805/) (2018)
- [Language Models are Few-Shot Learners](https://mlsanity.com/paper/2005.14165/) (2020)
- [Exploring the Limits of Transfer Learning with a Unified Text-to-Text Transformer](https://mlsanity.com/paper/1910.10683/) (2019)
- [RoBERTa: A Robustly Optimized BERT Pretraining Approach](https://mlsanity.com/paper/1907.11692/) (2019)
- [Retrieval-Augmented Generation for Knowledge-Intensive NLP Tasks](https://mlsanity.com/paper/2005.11401/) (2020)
- [BART: Denoising Sequence-to-Sequence Pre-training for Natural Language Generation, Translation, and Comprehension](https://mlsanity.com/paper/1910.13461/) (2019)
- [SQuAD: 100,000+ Questions for Machine Comprehension of Text](https://mlsanity.com/paper/1606.05250/) (2016)
- [Dense Passage Retrieval for Open-Domain Question Answering](https://mlsanity.com/paper/2004.04906/) (2020)
- [HotpotQA: A Dataset for Diverse, Explainable Multi-hop Question Answering](https://mlsanity.com/paper/1809.09600/) (2018)
- [TriviaQA: A Large Scale Distantly Supervised Challenge Dataset for Reading Comprehension](https://mlsanity.com/paper/1705.03551/) (2017)
- [REALM: Retrieval-Augmented Language Model Pre-Training](https://mlsanity.com/paper/2002.08909/) (2020)
- [Leveraging Passage Retrieval with Generative Models for Open Domain Question Answering](https://mlsanity.com/paper/2007.01282/) (2020)
- [DROP: A Reading Comprehension Benchmark Requiring Discrete Reasoning Over Paragraphs](https://mlsanity.com/paper/1903.00161/) (2019)
- [Reading Wikipedia to Answer Open-Domain Questions](https://mlsanity.com/paper/1704.00051/) (2017)
- [How Much Knowledge Can You Pack Into the Parameters of a Language Model?](https://mlsanity.com/paper/2002.08910/) (2020)
- [Latent Retrieval for Weakly Supervised Open Domain Question Answering](https://mlsanity.com/paper/1906.00300/) (2019)
- [UnifiedQA: Crossing Format Boundaries With a Single QA System](https://mlsanity.com/paper/2005.00700/) (2020)
- [The Web as a Knowledge-base for Answering Complex Questions](https://mlsanity.com/paper/1803.06643/) (2018)
- [TAPEX: Table Pre-training via Learning a Neural SQL Executor](https://mlsanity.com/paper/2107.07653/) (2021)
- [Answering Complex Open-Domain Questions with Multi-Hop Dense Retrieval](https://mlsanity.com/paper/2009.12756/) (2020)
- [Learning to Retrieve Reasoning Paths over Wikipedia Graph for Question Answering](https://mlsanity.com/paper/1911.10470/) (2019)
- [Break It Down: A Question Understanding Benchmark](https://mlsanity.com/paper/2001.11770/) (2020)
- [Multi-hop Reading Comprehension through Question Decomposition and Rescoring](https://mlsanity.com/paper/1906.02916/) (2019)
- [Unsupervised Question Decomposition for Question Answering](https://mlsanity.com/paper/2002.09758/) (2020)
- [Compositional Questions Do Not Necessitate Multi-hop Reasoning](https://mlsanity.com/paper/1906.02900/) (2019)
- [Cognitive Graph for Multi-Hop Reading Comprehension at Scale](https://mlsanity.com/paper/1905.05460/) (2019)
- [Dynamically Fused Graph Network for Multi-hop Reasoning](https://mlsanity.com/paper/1905.06933/) (2019)
- [Avoiding Reasoning Shortcuts: Adversarial Evaluation, Training, and Model Development for Multi-Hop QA](https://mlsanity.com/paper/1906.07132/) (2019)
- [Answering Complex Open-domain Questions Through Iterative Query Generation](https://mlsanity.com/paper/1910.07000/) (2019)
- [Understanding Dataset Design Choices for Multi-hop Reasoning](https://mlsanity.com/paper/1904.12106/) (2019)

## Built on by (most cited first, 7 of 7)

- [MPrompt: Exploring Multi-level Prompt Tuning for Machine Reading Comprehension](https://mlsanity.com/paper/2310.18167/) (2023)
- [POQD: Performance-Oriented Query Decomposer for Multi-vector retrieval](https://mlsanity.com/paper/2505.19189/) (2025)
- [Natural Response Generation for Chinese Reading Comprehension](https://mlsanity.com/paper/2302.08817/) (2023)
- [Multi-Hop Reasoning for Question Answering with Hyperbolic Representations](https://mlsanity.com/paper/2507.03612/) (2025)
- [Multi-Step Semantic Reasoning in Generative Retrieval](https://mlsanity.com/paper/2603.12368/) (2026)
- [Enhancing Pre-Trained Generative Language Models with Question Attended Span Extraction on Machine Reading Comprehension](https://mlsanity.com/paper/2404.17991/) (2024)
- [Incongruence Identification in Eyewitness Testimony](https://mlsanity.com/paper/2502.05650/) (2025)

## Abstract (arXiv)

Generative question answering (QA) models generate answers to questions either solely based on the parameters of the model (the closed-book setting) or additionally retrieving relevant evidence (the open-book setting). Generative QA models can answer some relatively complex questions, but the mechanism through which they do so is still poorly understood. We perform several studies aimed at better understanding the multi-hop reasoning capabilities of generative QA models. First, we decompose multi-hop questions into multiple corresponding single-hop questions, and find marked inconsistency in QA models' answers on these pairs of ostensibly identical question chains. Second, we find that models lack zero-shot multi-hop reasoning ability: when trained only on single-hop questions, models generalize poorly to multi-hop questions. Finally, we demonstrate that it is possible to improve models' zero-shot multi-hop reasoning capacity through two methods that approximate real multi-hop natural language (NL) questions by training on either concatenation of single-hop questions or logical forms (SPARQL). In sum, these results demonstrate that multi-hop reasoning does not emerge naturally in generative QA models, but can be encouraged by advances in training or modeling techniques.