
-
Can Small and Synthetic Benchmarks Drive Modeling Innovation? A Retrospective Study of Question Answering Modeling Approaches
Datasets are not only resources for training accurate, deployable system...
read it
-
Prefix-Tuning: Optimizing Continuous Prompts for Generation
Fine-tuning is the de facto way to leverage large pretrained language mo...
read it
-
WILDS: A Benchmark of in-the-Wild Distribution Shifts
Distribution shifts can cause significant degradation in a broad range o...
read it
-
In-N-Out: Pre-Training and Self-Training using Auxiliary Information for Out-of-Distribution Robustness
Consider a prediction setting where a few inputs (e.g., satellite images...
read it
-
Removing Spurious Features can Hurt Accuracy and Affect Groups Disproportionately
The presence of spurious features interferes with the goal of obtaining ...
read it
-
Beyond I.I.D.: Three Levels of Generalization for Question Answering on Knowledge Bases
Existing studies on question answering on knowledge bases (KBQA) mainly ...
read it
-
Selective Classification Can Magnify Disparities Across Groups
Selective classification, in which models are allowed to abstain on unce...
read it
-
Enabling certification of verification-agnostic networks via memory-efficient semidefinite programming
Convex relaxations have emerged as a promising approach for verifying de...
read it
-
RNNs can generate bounded hierarchical languages with optimal memory
Recurrent neural networks empirically generate natural language with hig...
read it
-
The EOS Decision and Length Extrapolation
Extrapolation to unseen sequence lengths is a challenge for neural gener...
read it
-
Learning Adaptive Language Interfaces through Decomposition
Our goal is to create an interactive natural language interface that eff...
read it
-
On the Importance of Adaptive Data Collection for Extremely Imbalanced Pairwise Tasks
Many pairwise classification tasks, such as paraphrase detection and ope...
read it
-
Task-Oriented Dialogue as Dataflow Synthesis
We describe an approach to task-oriented dialogue in which dialogue stat...
read it
-
Explore then Execute: Adapting without Rewards via Factorized Meta-Reinforcement Learning
We seek to efficiently learn by leveraging shared structure between diff...
read it
-
Robustness to Spurious Correlations via Human Annotations
The reliability of machine learning systems critically assumes that the ...
read it
-
Learning Abstract Models for Strategic Exploration and Fast Reward Transfer
Model-based reinforcement learning (RL) is appealing because (i) it enab...
read it
-
Concept Bottleneck Models
We seek to learn models that we can interact with using high-level conce...
read it
-
Simplifying Models with Unlabeled Output Data
We focus on prediction problems with high-dimensional outputs that are s...
read it
-
Selective Question Answering under Domain Shift
To avoid giving wrong answers, question answering (QA) models need to kn...
read it
-
Graph-based, Self-Supervised Program Repair from Diagnostic Feedback
We consider the problem of learning to repair programs from diagnostic f...
read it
-
Enabling Language Models to Fill in the Blanks
We present a simple approach for text infilling, the task of predicting ...
read it
-
An Investigation of Why Overparameterization Exacerbates Spurious Correlations
We study why overparameterization – increasing model size well beyond th...
read it
-
ExpBERT: Representation Engineering with Natural Language Explanations
Suppose we want to specify the inductive bias that married couples typic...
read it
-
Robust Encodings: A Framework for Combating Adversarial Typos
Despite excellent performance on many tasks, NLP systems are easily fool...
read it
-
Understanding Self-Training for Gradual Domain Adaptation
Machine learning systems must adapt to data distributions that evolve ov...
read it
-
Understanding and Mitigating the Tradeoff Between Robustness and Accuracy
Adversarial training augments the training set with perturbations to imp...
read it
-
Noise Induces Loss Discrepancy Across Groups for Linear Regression
We study the effect of feature noise (measurement error) on the discrepa...
read it
-
Distributionally Robust Neural Networks for Group Shifts: On the Importance of Regularization for Worst-Case Generalization
Overparameterized neural networks can be highly accurate on average on a...
read it
-
Learning Autocomplete Systems as a Communication Game
We study textual autocomplete—the task of predicting a full sentence fro...
read it
-
Shaping Visual Representations with Language for Few-shot Classification
Language is designed to convey useful information about the world, thus ...
read it
-
Verified Uncertainty Calibration
Applications such as weather forecasting and personalized medicine deman...
read it
-
Designing and Interpreting Probes with Control Tasks
Probes, supervised models trained to predict properties (like parts-of-s...
read it
-
Distributionally Robust Language Modeling
Language models are generally trained on data spanning a wide range of t...
read it
-
Certified Robustness to Adversarial Word Substitutions
State-of-the-art NLP models can often be fooled by adversaries that appl...
read it
-
A Tight Analysis of Greedy Yields Subexponential Time Approximation for Uniform Decision Tree
Decision Tree is a classic formulation of active learning: given n hypot...
read it
-
Selection Via Proxy: Efficient Data Selection For Deep Learning
Data selection methods such as active learning and core-set selection ar...
read it
-
Adversarial Training Can Hurt Generalization
While adversarial training can improve robust accuracy (against an adver...
read it
-
SPoC: Search-based Pseudocode to Code
We consider the task of mapping pseudocode to long programs that are fun...
read it
-
Maximum Weighted Loss Discrepancy
Though machine learning algorithms excel at minimizing the average loss ...
read it
-
Unlabeled Data Improves Adversarial Robustness
We demonstrate, theoretically and empirically, that adversarial robustne...
read it
-
On the Accuracy of Influence Functions for Measuring Group Effects
Influence functions estimate the effect of removing particular training ...
read it
-
Pre-training Graph Neural Networks
Many applications of machine learning in science and medicine, including...
read it
-
Pun Generation with Surprise
We tackle the problem of generating a pun sentence given a pair of homop...
read it
-
Unifying Human and Statistical Evaluation for Natural Language Generation
How can we measure whether a natural language generation system produces...
read it
-
Defending against Whitebox Adversarial Attacks via Randomized Discretization
Adversarial perturbations dramatically decrease the accuracy of state-of...
read it
-
Ambitious Data Science Can Be Painless
Modern data science research can involve massive computational experimen...
read it
-
Uncertainty Sampling is Preconditioned Stochastic Gradient Descent on Zero-One Loss
Uncertainty sampling, a popular active learning algorithm, is used to re...
read it
-
A Retrieve-and-Edit Framework for Predicting Structured Outputs
For the task of generating complex outputs such as source code, editing ...
read it
-
FrAngel: Component-Based Synthesis with Control Structures
In component-based program synthesis, the synthesizer generates a progra...
read it
-
Semidefinite relaxations for certifying robustness to adversarial examples
Despite their impressive performance on diverse tasks, neural networks f...
read it