Prompting Explicit and Implicit Knowledge for Multi-hop Question Answering Based on Human Reading Process

Cunjin Luo; Guangming Huang; Jiaxing Shen; Xia Sun; Yunfei Long

RECORD

Prompting Explicit and Implicit Knowledge for Multi-hop Question Answering Based on Human Reading Process

Guangming Huang, Yunfei Long, Cunjin Luo, Jiaxing Shen, Xia Sun

Computation and Language · Submitted 29 Feb 2024

abstract compiled · 2026-09-23 · no language-model review

The authors' main claim Experimental results show that PEI performs comparably to the state-of-the-art on HotpotQA.

how it is stated · asserted 1.00 read the record →

Comment (0) Read on arXiv ↗

Cite ▾

arXiv ID2402.19350 (version 6)

Pith IDZPPG7VCS, Pith's permanent record of this paper and its review

CitationCunjin Luo, Guangming Huang, Jiaxing Shen, Xia Sun, Yunfei Long. "Prompting Explicit and Implicit Knowledge for Multi-hop Question Answering Based on Human Reading Process." arXiv:2402.19350 (2024). https://pith.science/paper/ZPPG7VCS

Claim paper

The compiled record

The paper's own sentences, arranged. Each label is drawn from a closed vocabulary; the number beside it is the probability the classifier gave that label. Numbers in the tables are the authors', checked by code. No sentence in this record was written by a model.

Open the compiled record

Compiled by classifier-1.13.0 via TypeSafe API, question bank abstract-p1-20260923, 2026-09-23: 1 request, 6,779 input tokens.

What the abstract itself states, as probabilities: claims first 0.09 · states limitation 0.26 · human participants 0.07 · controlled comparison 0.48 · randomized assignment 0.01 · states quantitative result 0.06 · compares with prior results 0.93 · mentions code or data availability 0.02.

The abstract, sentence by sentence

arXiv listing abstract v6. Results and the main claim are underlined; show every role.

Pre-trained language models (PLMs) leverage chains-of-thought (CoT) to simulate human reasoning and inference processes, achieving proficient performance in multi-hop QA. However, a gap persists between PLMs' reasoning abilities and those of humans when tackling complex problems. Psychological studies suggest a vital connection between explicit information in passages and human prior knowledge during reading. Nevertheless, current research has given insufficient attention to linking input passages and PLMs' pre-training-based knowledge from the perspective of human cognition studies. In this study, we introduce a Prompting Explicit and Implicit knowledge (PEI) framework, which uses prompts to connect explicit and implicit knowledge, aligning with human reading process for multi-hop QA. We consider the input passages as explicit knowledge, employing them to elicit implicit knowledge through unified prompt reasoning. Furthermore, our model incorporates type-specific reasoning via prompts, a form of implicit knowledge. Experimental results show that PEI performs comparably to the state-of-the-art on HotpotQA. Ablation studies confirm the efficacy of our model in bridging and integrating explicit and implicit knowledge.

background contribution method result interpretation main claim

What the classifier reports for each sentence. The number is the probability it gave the label.
#	sentence	role	the authors' own work	states the main conclusion	strength
s0	Pre-trained language models (PLMs) leverage chains-of-thought (CoT) to simulate human reasoning and inference processes, achieving proficient performance in multi-hop QA.	background 1.00	0.07	0.04	asserted 1.00
s1	However, a gap persists between PLMs' reasoning abilities and those of humans when tackling complex problems.	gap 0.88	0.11	0.03	asserted 1.00
s2	Psychological studies suggest a vital connection between explicit information in passages and human prior knowledge during reading.	background 1.00	0.03	0.03	hedged possibility 0.90
s3	Nevertheless, current research has given insufficient attention to linking input passages and PLMs' pre-training-based knowledge from the perspective of human cognition studies.	gap 1.00	0.06	0.03	asserted 0.98
s4	In this study, we introduce a Prompting Explicit and Implicit knowledge (PEI) framework, which uses prompts to connect explicit and implicit knowledge, aligning with human reading process for multi-hop QA.	contribution 0.93	0.91	0.18	asserted 1.00
s5	We consider the input passages as explicit knowledge, employing them to elicit implicit knowledge through unified prompt reasoning.	method 0.99	0.84	0.07	asserted 1.00
s6	Furthermore, our model incorporates type-specific reasoning via prompts, a form of implicit knowledge.	method 0.92	0.86	0.05	asserted 1.00
s7	Experimental results show that PEI performs comparably to the state-of-the-art on HotpotQA.	result 1.00	0.96	0.38	asserted 1.00
s8	Ablation studies confirm the efficacy of our model in bridging and integrating explicit and implicit knowledge.	result 0.94	0.95	0.28	asserted 1.00

What kind of paper the classifier reports this is: method or tool 1.00 · evidence from observation or experiment 0.99.

What was examined, and what was not

sources read

arXiv listing abstract v6

not examined

full text
figures and tables
references and citing sentences
relations to other papers

levels run

P abstract

not run

S sections
T body sentences
C citations
R cross-paper relations

Every quoted sentence on this record is the authors'. Every label is from Pith's closed vocabulary, with the probability shown beside it. Every number in the measurements table is the authors', checked by code. Record as JSON

The paper

Original abstract

Pre-trained language models (PLMs) leverage chains-of-thought (CoT) to simulate human reasoning and inference processes, achieving proficient performance in multi-hop QA. However, a gap persists between PLMs' reasoning abilities and those of humans when tackling complex problems. Psychological studies suggest a vital connection between explicit information in passages and human prior knowledge during reading. Nevertheless, current research has given insufficient attention to linking input passages and PLMs' pre-training-based knowledge from the perspective of human cognition studies. In this study, we introduce a Prompting Explicit and Implicit knowledge (PEI) framework, which uses prompts to connect explicit and implicit knowledge, aligning with human reading process for multi-hop QA. We consider the input passages as explicit knowledge, employing them to elicit implicit knowledge through unified prompt reasoning. Furthermore, our model incorporates type-specific reasoning via prompts, a form of implicit knowledge. Experimental results show that PEI performs comparably to the state-of-the-art on HotpotQA. Ablation studies confirm the efficacy of our model in bridging and integrating explicit and implicit knowledge.

Human review

How Pith checks papers → · Full audit of this paper →

How the literature uses it

What this paper does with the arXiv papers it cites

Each citing sentence is taken from the citing paper's LaTeX source and typed by Pith's classifier. A label appears at probability 0.8 or more (0.95 for support), a contest at 0.7 or more; the quoted words are the citing authors'. How each label was checked · JSON

What it cites · 5 sentences without a confident label

Uses the method or tool of 2 papers

P-tuning v2 [ref] is employed to train soft prompts $P_t$ acquiring the weights of specific-type information of the given questions.

P-Tuning v2: Prompt Tuning Can Be Comparable to Fine-tuning Universally Across Scales and Tasks · Type Prompter · 1.00
ELECTRA-large [ref] serves as the foundation language model for both the type prompter and unified prompter modules.

ELECTRA: Pre-training Text Encoders as Discriminators Rather Than Generators · Implementation Details · 0.98

Compares with 1 paper

Illustrated in Table [ref], our proposed PEI model demonstrates superior performance across all evaluation metrics compared to all other baselines (except Beam Retrieval), and achieves comparable performance with Beam Retrieval [ref] on the HotpotQA dataset, highlighting the significant progress made by PEI for multi-hop QA.

End-to-End Beam Retrieval for Multi-Hop Question Answering · Main Results and Analysis · 1.00

Cites as background 4 papers

Some studies explored CoT prompt learning for multihop QA [ref] .

Interleaving Retrieval with Chain-of-Thought Reasoning for Knowledge-Intensive Multi-Step Questions · Chain-of-Thought Prompting · 0.93
Manual-CoT [ref] is designed to elicit the CoT reasoning ability through manual demonstrations.

Chain-of-Thought Prompting Elicits Reasoning in Large Language Models · Chain-of-Thought Prompting · 0.91
An automatic CoT prompting, AutoCoT [ref] , leverages diverse question sampling and reasoning chain generation to construct demonstrations, reducing the needs for human effort.

Automatic Chain of Thought Prompting in Large Language Models · Chain-of-Thought Prompting · 0.83

Discussion

Know something about this paper? Leave a comment, or highlight any passage above to comment on it.