Skip to main content ->
Ai2

Research - Papers

Explore a selection of our published work on a variety of key research challenges in AI.

Filter papers

ReadOnce Transformers: Reusable Representations of Text for Transformers

Shih-Ting LinAshish SabharwalTushar Khot
2021
ACL

While large-scale language models are extremely effective when directly fine-tuned on many end-tasks, such models learn to extract information and solve the task simultaneously from end-task… 

General-Purpose Question-Answering with Macaw

Oyvind TafjordPeter Clark
2021
arXiv

Despite the successes of pretrained language models, there are still few high-quality, general-purpose QA systems that are freely available. In response, we present MACAW, a versatile, generative… 

Did Aristotle Use a Laptop? A Question Answering Benchmark with Implicit Reasoning Strategies

Mor GevaDaniel KhashabiElad SegalJonathan Berant
2021
TACL

A key limitation in current datasets for multi-hop reasoning is that the required steps for answering the question are mentioned in it explicitly. In this work, we introduce STRATEGYQA, a question… 

ProofWriter: Generating Implications, Proofs, and Abductive Statements over Natural Language

Oyvind TafjordB. D. MishraP. Clark
2021
Findings of ACL

Transformers have been shown to emulate logical deduction over natural language theories (logical rules expressed in natural language), reliably assigning true/false labels to candidate… 

ParsiNLU: A Suite of Language Understanding Challenges for Persian

Daniel KhashabiArman CohanSiamak Shakeriet al.
2021
TACL

Despite the progress made in recent years in addressing natural language understanding (NLU) challenges, the majority of this progress remains to be concentrated on resource-rich languages like… 

Critical Thinking for Language Models

Gregor BetzChristian VoigtKyle Richardson
2021
IWCS

This paper takes a first step towards a critical thinking curriculum for neural auto-regressive language models. We introduce a synthetic text corpus of deductively valid arguments, and use this… 

Temporal Reasoning on Implicit Events from Distant Supervision

Ben ZhouKyle RichardsonQiang NingD. Roth
2021
NAACL

Existing works on temporal reasoning among events described in text focus on modeling relationships between explicitly mentioned events and do not handle event end time effectively. However, human… 

Text Modular Networks: Learning to Decompose Tasks in the Language of Existing Models

Tushar KhotDaniel KhashabiKyle RichardsonAshish Sabharwal
2021
NAACL

A common approach to solve complex tasks is by breaking them down into simple sub-problems that can then be solved by simpler modules. However, these approaches often need to be designed and trained… 

Thinking Aloud: Dynamic Context Generation Improves Zero-Shot Reasoning Performance of GPT-2

G. BetzKyle RichardsonChristian Voigt
2021
arXiv

Thinking aloud is an effective meta-cognitive strategy human reasoners apply to solve difficult problems. We suggest to improve the reasoning ability of pre-trained neural language models in a… 

Information to Wisdom: Commonsense Knowledge Extraction and Compilation

Simon RazniewskiNiket TandonAparna S. Varde
2021
WSDM '21: Proceedings of the 14th ACM International Conference on Web Search and Data Mining

Commonsense knowledge is a foundational cornerstone of artificial intelligence applications. Whereas information extraction and knowledge base construction for instance-oriented assertions, such as…