Research

Papers + Code

Peer-review is the lifeblood of scientific validation and a guardrail against runaway hype in AI. Our commitment to publishing in the top venues reflects our grounding in what is real, reproducible, and truly innovative.

Sign-OPT: Defending the hard-label black-box cyber attack

SenSR: the first practical algorithm for individual fairness

Why Gradient Clipping accelerates training for neural networks

CLEVRER: The first video dataset for neuro-symbolic reasoning

Fast and efficient black-box testing for AI cybersecurity

Learning to learn with distributional signatures for text data

Layer-wise federated learning with FedMA

Learning Rate Rewinding for elegant neural network pruning

Generative Ratio Matching Networks

Implementation Matters in Deep RL: A Case Study on PPO and TRPO

Once for All: Train One Network and Specialize it for Efficient Deployment

Deep Audio Priors Emerge From Harmonic Convolutional Networks

A Closer Look at Deep Policy Gradients

Deep Symbolic Superoptimization Without Human Knowledge

Reshaping Diverse Planning

Adversarial Robustness vs Model Compression, or Both?

Point-Voxel CNN for Efficient 3D Deep Learning

CASTER: An AI framework for preventing adverse reactions to medication

On the Design of Black-box Adversarial Examples by Leveraging Gradient-free Optimization and Operator Splitting Method

Reasoning about Human-Object Interactions through Dual Attention Networks

LaSO: Label-Set Operations networks for multi-label few-shot learning

SpotTune: Transfer Learning through Adaptive Fine-tuning

RepMet: Representative-based metric learning for classification and one-shot object detection

ObjectNet: A bias-controlled dataset object recognition

Sentence Embedding Alignment for Lifelong Relation Extraction

Moments in Time Dataset: one million videos for event understanding

Anti-Money Laundering in Bitcoin: Experimenting with Graph Convolutional Networks for Financial Forensics

Self-supervised Moving Vehicle Tracking with Stereo Sound

The sound of motions

TSM: Temporal Shift Module for Efficient Video Understanding

Graph Convolutional Networks for Temporal Action Localization

On the Convergence of A Class of Adam-Type Algorithms for Non-Convex Optimization

Online AI planning with graph neural networks and adaptive scheduling

Reading between the lines with graph deep learning for NLP

A unifying framework for expectation-aware AI planning

A content-aware attack generator for AI cybersecurity

Automating machine learning with a joint selection framework

EvolveGCN: Evolving Graph Convolutional Networks for Dynamic Graphs

Bio-Inspired Hashing for Unsupervised Similarity Search

Embedding Compression with Isotropic Iterative Quantization

Scalable Spike Source Localization in Extracellular Recordings using Amortized Variational Inference

ZO-AdaMM: Derivative-free optimization for black-box problems

Big-Little-Video-Net: Work smarter, not harder, for video understanding

Topics are more meaningful than words. AI for comparative literature.

Imitation learning from observations

Using geometry to understand documents

SPAHM: Parameter matching for model fusion

Alleviating label switching with optimal transport

Visual Concept-Metaconcept Learning

Causal inference is expensive. Here’s an algorithm for fixing that.

Class-wise rationalization: teaching AI to weigh pros and cons

New tricks from old dogs: multi-source transfer learning

Reverse-engineering causal graphs with soft interventions

SimVAE: Simulator-Assisted Training for Interpretable Generative Models

Cross-channel Communication Networks

The Future of Work: How New Technologies Are Transforming Tasks

Learning to Learn without Forgetting By Maximizing Transfer and Minimizing Interference

Seeing What a GAN Cannot Generate

Watch, Reason and Code: Learning to Represent Videos Using Program

Facial Image-to-Video Translation by a Hidden Affine Transformation

Defensive Quantization: When Efficiency Meets Robustness

Local Unsupervised Learning for Image Analysis

Neural language models as psycholinguistic subjects: Representations of syntactic state

Beyond RNNs: Positional Self-Attention with Co-Attention for Video Question Answering

Deep Leakage from Gradients

Write, Execute, Assess: Program Synthesis with a REPL

Variational Russian Roulette for Deep Bayesian Nonparametrics

Imposing Label-Relational Inductive Bias for Extremely Fine-Grained Entity Typing

Bayesian Nonparametric Federated Learning of Neural Networks

The Neuro-Symbolic Concept Learner: Interpreting Scenes, Words, and Sentences From Natural Supervision

Unsupervised learning by competing hidden units

Revisiting RCNN: On Awakening the Classification Power of Faster RCNN

Structural Supervision Improves Learning of Non-Local Grammatical Dependencies

Tight Certificates of Adversarial Robustness

Evaluating the Robustness of Neural Networks: An Extreme Value Theory Approach

Weakly Supervised Dense Event Captioning in Videos

Neural Network Robustness Certification with General Activation Functions

Scalable Graph Learning for Anti-Money Laundering: A First Look

CNN-Cert: An Efficient Framework for Certifying Robustness of Convolutional Neural Networks

GAN Dissection: Visualizing and Understanding Generative Adversarial Networks

Collective Online Learning of Gaussian Processes in Massive Multi-Agent Systems

Unsupervised learning with contrastive latent variable models

Co-regularized Alignment for Unsupervised Domain Adaptation

StNet: Local and Global Spatial-Temporal Modeling for Action Recognition

Experimental Design for Cost-Aware Learning of Causal Graphs

MiME: Multilevel Medical Embedding of Electronic Health Records for Predictive Healthcare

Neural-Symbolic VQA: Disentangling Reasoning from Vision and Language Understanding

signSGD via Zeroth-Order Oracle

Learning Entropic Wasserstein Embeddings

Constrained Generation of Semantically Valid Graphs via Regularizing Variational Autoencoders

emrQA: A Large Corpus for Question Answering on Electronic Medical Records

Deriving Machine Attention from Human Rationales

Controllable Image-to-Video Translation: A Case Study on Facial Expression Generation

Structured Adversarial Attack: Towards General Implementation and Better Interpretability

Delta-encoder: an effective sample synthesis method for few-shot object recognition

AutoZOOM: Autoencoder-based Zeroth Order Optimization Method for Attacking Black-box Neural Networks

Zeroth-Order Stochastic Variance Reduction for Nonconvex Optimization

Jointly Optimize Data Augmentation and Network Training: Adversarial Data Augmentation in Human Pose Estimation

Dialog-based Interactive Image Retrieval

Learning to Teach in Cooperative Multiagent Reinforcement Learning

The Sound of Pixels

Learning to Separate Object Sounds by Watching Unlabeled Video

The Lottery Ticket Hypothesis: Finding Sparse, Trainable Neural Networks

Attacking Visual Language Grounding with Adversarial Examples: A Case Study on Neural Image Captioning

BlockDrop: Dynamic Inference Paths in Residual Networks

Action Centered Contextual Bandits

Domain Adaptation by Using Causal Inference to Predict Invariant Conditional Distributions

Gradient Descent for Spiking Neural Networks

Ensemble Estimation of Information Divergence