paper-with-me

Papers AI2 Reasoning Challenge

“AI2 Reasoning Challenge” 태그가 달린 논문 10편 · 필터 해제

Towards a Mechanistic Interpretation of Multi-Step Reasoning Capabilities of Language Models

2023-10-23 · Yifan Hou, Jiaoda Li, Yu Fei, Alessandro Stolfo 외

Recent work has shown that language models (LMs) have strong multi-step (i.e., procedural) reasoning capabilities. However, it is unclear whether LMs perform these tasks by cheating with answers memorized from pretrainin…

AI2 Reasoning Challenge

Think you have Solved Direct-Answer Question Answering? Try ARC-DA, the Direct-Answer AI2 Reasoning Challenge

2021-02-05 · Sumithra Bhakthavatsalam, Daniel Khashabi, Tushar Khot, Bhavana Dalvi Mishra 외

We present the ARC-DA dataset, a direct-answer ("open response", "freeform") version of the ARC (AI2 Reasoning Challenge) multiple-choice dataset. While ARC has been influential in the community, its multiple-choice form…

AI2 Reasoning ChallengeARCMultiple-choiceNatural Questions+2

Challenge Closed-book Science Exam: A Meta-learning Based Question Answering System

2020-04-26 · Xinyue Zheng, Peng Wang, Qigang Wang, Zhongchao shi

Prior work in standardized science exams requires support from large text corpus, such as targeted science corpus fromWikipedia or SimpleWikipedia. However, retrieving knowledge from the large corpus is time-consuming an…

AI2 Reasoning ChallengeARCLanguage ModelingLanguage Modelling+4

Team SVMrank: Leveraging Feature-rich Support Vector Machines for Ranking Explanations to Elementary Science Questions

2019-11-01 · WS 2019 11 · Jennifer D{'}Souza, Isaiah On Mulang{'}, o, S{\"o}ren Auer

The TextGraphs 2019 Shared Task on Multi-Hop Inference for Explanation Regeneration (MIER-19) tackles explanation generation for answers to elementary science questions. It builds on the AI2 Reasoning Challenge 2018 (ARC…

AI2 Reasoning ChallengeARCExplanation GenerationLearning-To-Rank+2

Alignment over Heterogeneous Embeddings for Question Answering

2019-06-01 · NAACL 2019 6 · Vikas Yadav, Steven Bethard, Mihai Surdeanu

We propose a simple, fast, and mostly-unsupervised approach for non-factoid question answering (QA) called Alignment over Heterogeneous Embeddings (AHE). AHE simply aligns each word in the question and candidate answer w…

AI2 Reasoning ChallengeARCQuestion AnsweringSentence+1

An Interface for Annotating Science Questions

2018-11-01 · EMNLP 2018 11 · Michael Boratko, Harshit Padigela, Divyendra Mikkilineni, Pritish Yuvraj 외

Recent work introduces the AI2 Reasoning Challenge (ARC) and the associated ARC dataset that partitions open domain, complex science questions into an Easy Set and a Challenge Set. That work includes an analysis of 100 q…

AI2 Reasoning ChallengeARCGeneral ClassificationInformation Retrieval

Learning to Attend On Essential Terms: An Enhanced Retriever-Reader Model for Open-domain Question Answering

2018-08-28 · NAACL 2019 6 · Jianmo Ni, Chenguang Zhu, Weizhu Chen, Julian McAuley

Open-domain question answering remains a challenging task as it requires models that are capable of understanding questions and answers, collecting useful information, and reasoning over evidence. Previous work typically…

AI2 Reasoning ChallengeARCMultiple-choiceMultiple Choice Question Answering (MCQA)+3

A Systematic Classification of Knowledge, Reasoning, and Context within the ARC Dataset

2018-06-01 · WS 2018 7 · Michael Boratko, Harshit Padigela, Divyendra Mikkilineni, Pritish Yuvraj 외

The recent work of Clark et al. introduces the AI2 Reasoning Challenge (ARC) and the associated ARC dataset that partitions open domain, complex science questions into an Easy Set and a Challenge Set. That paper includes…

AI2 Reasoning ChallengeARCGeneral ClassificationInformation Retrieval+2

KG^2: Learning to Reason Science Exam Questions with Contextual Knowledge Graph Embeddings

2018-05-31 · Yuyu Zhang, Hanjun Dai, Kamil Toraman, Le Song

The AI2 Reasoning Challenge (ARC), a new benchmark dataset for question answering (QA) has been recently released. ARC only contains natural science questions authored for human exams, which are hard to answer and requir…

AI2 Reasoning ChallengeARCKnowledge Graph EmbeddingsKnowledge Graphs+1

Think you have Solved Question Answering? Try ARC, the AI2 Reasoning Challenge

2018-03-14 · Peter Clark, Isaac Cowhey, Oren Etzioni, Tushar Khot 외

We present a new question set, text corpus, and baselines assembled to encourage AI research in advanced question answering. Together, these constitute the AI2 Reasoning Challenge (ARC), which requires far more powerful …

AI2 Reasoning ChallengeARCQuestion AnsweringRetrieval
1–10 / 10