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

Better Retrieval May Not Lead to Better Question Answering

Zhengzhong LiangTushar KhotSteven BethardAshish Sabharwal
2022
arXiv

Considerable progress has been made recently in open-domain question answering (QA) problems, which require Information Retrieval (IR) and Reading Comprehension (RC). A popular approach to improve… 

Saturated Transformers are Constant-Depth Threshold Circuits

William MerrillAshish SabharwalNoah A. Smith
2022
TACL

Transformers have become a standard neural network architecture for many NLP problems, motivating theoretical analysis of their power in terms of formal languages. Recent work has shown that… 

Memory-assisted prompt editing to improve GPT-3 after deployment

Aman MadaanNiket TandonPeter ClarkYiming Yang
2022
ACL • Workshop on Commonsense Reasoning

Large LMs such as GPT-3 are powerful, but can commit mistakes that are obvious to humans. For example, GPT-3 would mistakenly interpret "What word is similar to good?" to mean a homonym, while the… 

Multi-Modal Answer Validation for Knowledge-Based VQA

Jialin WuJiasen LuAshish SabharwalR. Mottaghi
2022
AAAI

The problem of knowledge-based visual question answering involves answering questions that require external knowledge in addition to the content of the image. Such knowledge typically comes in a… 

Pushing the Limits of Rule Reasoning in Transformers through Natural Language Satisfiability

Kyle RichardsonAshish Sabharwal
2022
AAAI

Investigating the reasoning abilities of transformer models, and discovering new challenging tasks for them, has been a topic of much interest. Recent studies have found these models to be… 

MuSiQue: Multihop Questions via Single-hop Question Composition

Harsh TrivediNiranjan BalasubramanianTushar KhotAshish Sabharwal
2022
TACL

Multihop reasoning remains an elusive goal as existing multihop benchmarks are known to be largely solvable via shortcuts. Can we create a question answering (QA) dataset that, by construction,… 

PROMPT WAYWARDNESS: The Curious Case of Discretized Interpretation of Continuous Prompts

Daniel KhashabiShan LyuSewon MinYejin Choi
2022
NAACL

Fine-tuning continuous prompts for target tasks has recently emerged as a compact alternative to full model fine-tuning. Motivated by these promising results, we investigate the feasibility of… 

DREAM: Improving Situational QA by First Elaborating the Situation

Yuling GuBhavana Dalvi MishraPeter Clark
2021
NAACL

When people answer questions about a specific situation, e.g., "I cheated on my mid-term exam last week. Was that wrong?", cognitive science suggests that they form a mental picture of that… 

BeliefBank: Adding Memory to a Pre-Trained Language Model for a Systematic Notion of Belief

Nora KassnerOyvind TafjordH. SchutzeP. Clark
2021
EMNLP

Although pretrained language models (PTLMs) have been shown to contain significant amounts of world knowledge, they can still produce inconsistent answers to questions when probed, even after using… 

Explaining Answers with Entailment Trees

Bhavana DalviPeter A. JansenOyvind TafjordPeter Clark
2021
EMNLP

Our goal, in the context of open-domain textual question-answering (QA), is to explain answers by not just listing supporting textual evidence (“rationales”), but also showing how such evidence…