Search Results for author:
Found 71 papers, 30 papers with code
Distantly-Supervised Dense Retrieval Enables Open-Domain Question Answering without Evidence Annotation
This paper investigates whether models can learn to find evidence from a large corpus, with only distant supervision from answer labels for model training, thereby generating no additional annotation cost.
Toward Gender-Inclusive Coreference Resolution: An Analysis of Gender and Bias Throughout the Machine Learning Lifecycle*
Such inferences raise the risk of systematic biases in coreference resolution systems, including biases that can harm binary and non-binary trans and cis stakeholders.
HateCOT: An Explanation-Enhanced Dataset for Generalizable Offensive Speech Detection via Large Language Models
We show that pre-training models for the detection of offensive content on HateCOT significantly boots open-sourced Language Models on three benchmark datasets in both zero and few-shot settings, despite differences in domain and task.}
Successfully Guiding Humans with Imperfect Instructions by Highlighting Potential Errors and Suggesting Corrections
This paper addresses the challenge of leveraging imperfect language models to guide human decision-making in the context of a grounded navigation task.
PRISE: Learning Temporal Action Abstractions as a Sequence Compression Problem
To do so, we bring a subtle but critical component of LLM training pipelines -- input tokenization via byte pair encoding (BPE) -- to the seemingly distant task of learning skills of variable time span in continuous control domains.
Premier-TACO is a Few-Shot Policy Learner: Pretraining Multitask Representation via Temporal Action-Driven Contrastive Loss
We present Premier-TACO, a multitask feature representation learning approach designed to improve few-shot policy learning efficiency in sequential decision-making tasks.
Toxicity Detection is NOT all you Need: Measuring the Gaps to Supporting Volunteer Content Moderators
Extensive efforts in automated approaches for content moderation have been focused on developing models to identify toxic, offensive, and hateful content with the aim of lightening the load for moderators.
DrM: Mastering Visual Reinforcement Learning through Dormant Ratio Minimization
To quantify this inactivity, we adopt dormant ratio as a metric to measure inactivity in the RL agent's network.
Hallucination Detection for Grounded Instruction Generation
We investigate the problem of generating instructions to guide humans to navigate in simulated residential environments.
Towards Conceptualization of "Fair Explanation": Disparate Impacts of anti-Asian Hate Speech Explanations on Content Moderators
We apply this method in the context of content moderation of potential hate speech, and its differential impact on Asian vs. non-Asian proxy moderators, across explanation approaches (saliency map and counterfactual explanation).
Large Language Models Help Humans Verify Truthfulness -- Except When They Are Convincingly Wrong
To reduce over-reliance on LLMs, we ask LLMs to provide contrastive information - explain both why the claim is true and false, and then we present both sides of the explanation to users.
Progressively Efficient Learning
By equipping a learning agent with an abstract, dynamic language and an intrinsic motivation to learn with minimal communication effort, CEIL leads to emergence of a human-like pattern where the learner and the teacher communicate progressively efficiently by exchanging increasingly more abstract intentions.
The Impact of Explanations on Fairness in Human-AI Decision-Making: Protected vs Proxy Features
In this work, we study the effect of the presence of protected and proxy features on participants' perception of model fairness and their ability to improve demographic parity over an AI alone.
FairPrism: Evaluating Fairness-Related Harms in Text Generation
It is critical to measure and mitigate fairness- related harms caused by AI text generation systems, including stereotyping and demeaning harms.
TACO: Temporal Latent Action-Driven Contrastive Loss for Visual Reinforcement Learning
Despite recent progress in reinforcement learning (RL) from raw pixel data, sample inefficiency continues to present a substantial obstacle.
Evaluating the Social Impact of Generative AI Systems in Systems and Society
We move toward a standard approach in evaluating a generative AI system for any modality, in two overarching categories: what is able to be evaluated in a base system that has no predetermined application and what is able to be evaluated in society.
What Else Do I Need to Know? The Effect of Background Information on Users' Reliance on QA Systems
In this work, we study how users interact with QA systems in the absence of sufficient information to assess their predictions.
It Takes Two to Tango: Navigating Conceptualizations of NLP Tasks and Measurements of Performance
Progress in NLP is increasingly measured through benchmarks; hence, contextualizing progress requires understanding when and why practitioners may disagree about the validity of benchmarks.
Define, Evaluate, and Improve Task-Oriented Cognitive Capabilities for Instruction Generation Models
Recent work studies the cognitive capabilities of language models through psychological tests designed for humans.
How do Authors' Perceptions of their Papers Compare with Co-authors' Perceptions and Peer-review Decisions?
In a top-tier computer science conference (NeurIPS 2021) with more than 23, 000 submitting authors and 9, 000 submitted papers, we survey the authors on three questions: (i) their predicted probability of acceptance for each of their papers, (ii) their perceived ranking of their own papers based on scientific contribution, and (iii) the change in their perception about their own papers after seeing the reviews.
What's Different between Visual Question Answering for Machine "Understanding" Versus for Accessibility?
We aim to answer this question by evaluating discrepancies between machine "understanding" datasets (VQA-v2) and accessibility datasets (VizWiz) by evaluating a variety of VQA models.
Theory-Grounded Measurement of U.S. Social Stereotypes in English Language Models
NLP models trained on text have been shown to reproduce human stereotypes, which can magnify harms to marginalized groups when systems are deployed at scale.
Deconstructing NLG Evaluation: Evaluation Practices, Assumptions, and Their Implications
There are many ways to express similar things in text, which makes evaluating natural language generation (NLG) systems difficult.
A Framework for Learning to Request Rich and Contextually Useful Information from Humans
We show that the agent can take advantage of different types of information depending on the context, and analyze the benefits and challenges of learning the assistance-requesting policy when the assistant can recursively decompose tasks into subtasks.
Distantly-Supervised Evidence Retrieval Enables Question Answering without Evidence Annotation
Open-domain question answering answers a question based on evidence retrieved from a large corpus.
Learning When and What to Ask: a Hierarchical Reinforcement Learning Framework
Results on a simulated human-assisted navigation problem demonstrate the effectiveness of our framework: aided with an interaction policy learned by our method, a navigation policy achieves up to a 7× improvement in task success rate compared to performing tasks only by itself.
Hierarchical Reinforcement Learning
reinforcement-learning
+1
From Human Explanation to Model Interpretability: A Framework Based on Weight of Evidence
We take inspiration from the study of human explanation to inform the design and evaluation of interpretability methods in machine learning.
BIG-bench Machine Learning
Interpretable Machine Learning
+1
Multi-Step Reasoning Over Unstructured Text with Beam Dense Retrieval
Complex question answering often requires finding a reasoning chain that consists of multiple evidence pieces.
A Large Scale Randomized Controlled Trial on Herding in Peer-Review Discussions
Peer review is the backbone of academia and humans constitute a cornerstone of this process, being responsible for reviewing papers and making the final acceptance/rejection decisions.
A Novice-Reviewer Experiment to Address Scarcity of Qualified Reviewers in Large Conferences
Conference peer review constitutes a human-computation process whose importance cannot be overstated: not only it identifies the best submissions for acceptance, but, ultimately, it impacts the future of the whole research area by promoting some ideas and restraining others.
Prior and Prejudice: The Novice Reviewers' Bias against Resubmissions in Conference Peer Review
Modern machine learning and computer science conferences are experiencing a surge in the number of submissions that challenges the quality of peer review as the number of competent reviewers is growing at a much slower rate.
On the Potential of Lexico-logical Alignments for Semantic Parsing to SQL Queries
Large-scale semantic parsing datasets annotated with logical forms have enabled major advances in supervised approaches.
Active Imitation Learning from Multiple Non-Deterministic Teachers: Formulation, Challenges, and Algorithms
We formulate the problem of learning to imitate multiple, non-deterministic teachers with minimal interaction cost.
Operationalizing the Legal Principle of Data Minimization for Personalization
Article 5(1)(c) of the European Union's General Data Protection Regulation (GDPR) requires that "personal data shall be [...] adequate, relevant, and limited to what is necessary in relation to the purposes for which they are processed (`data minimisation')".
Language (Technology) is Power: A Critical Survey of "Bias" in NLP
We survey 146 papers analyzing "bias" in NLP systems, finding that their motivations are often vague, inconsistent, and lacking in normative reasoning, despite the fact that analyzing "bias" is an inherently normative process.
Active Imitation Learning with Noisy Guidance
Imitation learning algorithms provide state-of-the-art results on many structured prediction tasks by learning near-optimal search policies.
Towards Automatic Generation of Questions from Long Answers
Therefore, we propose a novel evaluation benchmark to assess the performance of existing AQG systems for long-text answers.
Meta-Learning for Few-Shot NMT Adaptation
We frame the adaptation of NMT systems as a meta-learning problem, where we learn to adapt to new unseen domains based on simulated offline meta-training domain adaptation tasks.
Toward Gender-Inclusive Coreference Resolution
Correctly resolving textual mentions of people fundamentally entails making inferences about those people.
Weight of Evidence as a Basis for Human-Oriented Explanations
Interpretability is an elusive but highly sought-after characteristic of modern machine learning methods.
Global Voices: Crossing Borders in Automatic News Summarization
We construct Global Voices, a multilingual dataset for evaluating cross-lingual summarization methods.
Learning Effective Exploration Strategies For Contextual Bandits
We develop a meta-learning algorithm, MELEE, that learns an exploration policy based on simulated, synthetic contextual bandit tasks.
Help, Anna! Visual Navigation with Natural Multimodal Assistance via Retrospective Curiosity-Encouraging Imitation Learning
An agent solving tasks in a HANNA environment can leverage simulated human assistants, called ANNA (Automatic Natural Navigation Assistants), which, upon request, provide natural language and visual instructions to direct the agent towards the goals.
Reinforcement Learning with Convex Constraints
In standard reinforcement learning (RL), a learning agent seeks to optimize the overall reward.
Answer-based Adversarial Training for Generating Clarification Questions
We present an approach for generating clarification questions with the goal of eliciting new information that would make the given textual context more complete.
Non-Monotonic Sequential Text Generation
Standard sequential generation methods assume a pre-specified generation order, such as text generation methods which generate words from left to right.
Meta-Learning for Contextual Bandit Exploration
We describe MELEE, a meta-learning algorithm for learning a good exploration policy in the interactive contextual bandit setting.
Warm-starting Contextual Bandits: Robustly Combining Supervised and Bandit Feedback
We investigate the feasibility of learning from a mix of both fully-labeled supervised data and contextual bandit data.
Improving fairness in machine learning systems: What do industry practitioners need?
The potential for machine learning (ML) systems to amplify social inequities and unfairness is receiving increasing popular and academic attention.
Contextual Memory Trees
We design and study a Contextual Memory Tree (CMT), a learning memory controller that inserts new memories into an experience store of unbounded size.
Learning to Ask Good Questions: Ranking Clarification Questions using Neural Expected Value of Perfect Information
Inquiry is fundamental to communication, and machines cannot effectively collaborate with humans unless they can ask questions.
Datasheets for Datasets
The machine learning community currently has no standardized process for documenting datasets, which can lead to severe consequences in high-stakes domains.
Technical Report: When Does Machine Learning FAIL? Generalized Transferability for Evasion and Poisoning Attacks
Our model allows us to consider a wide range of weaker adversaries who have limited control and incomplete knowledge of the features, learning algorithms and training instances utilized.
Hierarchical Imitation and Reinforcement Learning
We study how to effectively leverage expert feedback to learn sequential decision-making policies.
Residual Loss Prediction: Reinforcement Learning With No Incremental Feedback
We consider reinforcement learning and bandit structured prediction problems with very sparse loss feedback: only at the end of an episode.
Towards Linguistically Generalizable NLP Systems: A Workshop and Shared Task
This paper presents a summary of the first Workshop on Building Linguistically Generalizable Natural Language Processing Systems, and the associated Build It Break It, The Language Edition shared task.
The UMD Neural Machine Translation Systems at WMT17 Bandit Learning Task
We describe the University of Maryland machine translation systems submitted to the WMT17 German-English Bandit Learning Task.
Reinforcement Learning for Bandit Neural Machine Translation with Simulated Human Feedback
Machine translation is a natural candidate problem for reinforcement learning from human feedback: users provide quick, dirty ratings on candidate translations to guide a system to improve.
The Amazing Mysteries of the Gutter: Drawing Inferences Between Panels in Comic Book Narratives
While computers can now describe what is explicitly depicted in natural images, in this paper we examine whether they can understand the closure-driven narratives conveyed by stylized artwork and dialogue in comic book panels.
Opponent Modeling in Deep Reinforcement Learning
Opponent modeling is necessary in multi-agent settings where secondary agents with competing goals also adapt their strategies, yet it remains challenging because strategies interact with each other and change.
On Correcting Inputs: Inverse Optimization for Online Structured Prediction
Algorithm designers typically assume that the input data is correct, and then proceed to find "optimal" or "sub-optimal" solutions using this input data.
Learning to Search for Dependencies
We demonstrate that a dependency parser can be built using a credit assignment compiler which removes the burden of worrying about low-level machine learning details from the parser implementation.
Learning Reductions that Really Work
We provide a summary of the mathematical and computational techniques that have enabled learning reductions to effectively address a wide class of problems, and show that this approach to solving machine learning problems can be broadly useful.
Learning to Search Better Than Your Teacher
Methods for learning to search for structured prediction typically imitate a reference policy, with existing theoretical guarantees demonstrating low regret compared to that reference.
A Credit Assignment Compiler for Joint Prediction
Many machine learning applications involve jointly predicting multiple mutually dependent output variables.
Frustratingly Easy Domain Adaptation
We describe an approach to domain adaptation that is appropriate exactly in the case when one has enough ``target'' data to do slightly better than just using only ``source'' data.
Search-based Structured Prediction
We present Searn, an algorithm for integrating search and learning to solve complex structured prediction problems such as those that occur in natural language, speech, computational biology, and vision.
A Bayesian Model for Discovering Typological Implications
A standard form of analysis for linguistic typology is the universal implication.
