Search Results for author:
Found 136 papers, 80 papers with code
Predicting deliberative outcomes
Our games take as input, e. g., UN resolution to be voted on, and map such contexts to initial strategies, player utilities, and interactions.
Composing Molecules with Multiple Property Constraints
These rationales are identified from molecules as substructures that are likely responsible for each property of interest.
Predicting deliberative outcomes
Our games take as input, e. g., UN resolution to be voted on, and map such contexts to initial strategies, player utilities, and interactions.
ProxelGen: Generating Proteins as 3D Densities
We develop ProxelGen, a protein structure generative model that operates on 3D densities as opposed to the prevailing 3D point cloud representations.
Thought calibration: Efficient and confident test-time scaling
Reasoning large language models achieve impressive test-time scaling by thinking for longer, but this performance gain comes at significant compute cost.
Protein FID: Improved Evaluation of Protein Structure Generative Models
Protein structure generative models have seen a recent surge of interest, but meaningfully evaluating them computationally is an active area of research.
Hierarchical protein backbone generation with latent and structure diffusion
We propose a hierarchical protein backbone generative model that separates coarse and fine-grained details.
ProtComposer: Compositional Protein Structure Generation with 3D Ellipsoids
We develop ProtComposer to generate protein structures conditioned on spatial protein layouts that are specified via a set of 3D ellipsoids capturing substructure shapes and semantics.
LEAPS: A discrete neural sampler via locally equivariant networks
This property allows us to propose a scalable training algorithm for the rate matrix such that the variance of the importance weights associated to the CTMC are minimal.
Inference-Time Scaling for Diffusion Models beyond Scaling Denoising Steps
In this work, we explore the inference-time scaling behavior of diffusion models beyond increasing denoising steps and investigate how the generation performance can further improve with increased computation.
Scaling Inference Time Compute for Diffusion Models
In this work, we explore the inference-time scaling behavior of diffusion models beyond increasing denoising steps and investigate how the generation performance can further improve with increased computation.
An Information Criterion for Controlled Disentanglement of Multimodal Data
Multimodal representation learning seeks to relate and decompose information inherent in multiple modalities.
A Cosmic-Scale Benchmark for Symmetry-Preserving Data Processing
Efficiently processing structured point cloud data while preserving multiscale information is a key challenge across domains, from graphics to atomistic modeling.
Hamiltonian Score Matching and Generative Flows
Classical Hamiltonian mechanics has been widely used in machine learning in the form of Hamiltonian Monte Carlo for applications with predetermined force fields.
Generator Matching: Generative modeling with arbitrary Markov processes
We introduce generator matching, a modality-agnostic framework for generative modeling using arbitrary Markov processes.
Fictitious Synthetic Data Can Improve LLM Factuality via Prerequisite Learning
Recent studies have identified one aggravating factor of LLM hallucinations as the knowledge inconsistency between pre-training and fine-tuning, where unfamiliar fine-tuning data mislead the LLM to fabricate plausible but wrong outputs.
Fine-Tuning Discrete Diffusion Models via Reward Optimization with Applications to DNA and Protein Design
Finally, we demonstrate the effectiveness of DRAKES in generating DNA and protein sequences that optimize enhancer activity and protein stability, respectively, important tasks for gene therapies and protein-based therapeutics.
Think While You Generate: Discrete Diffusion with Planned Denoising
DDPD outperforms traditional denoiser-only mask diffusion methods, achieving superior results on language modeling benchmarks such as text8, OpenWebText, and token-based image generation on ImageNet $256 \times 256$.
Identifying biological perturbation targets through causal differential networks
Given a pair of observational and interventional datasets, the goal is to isolate the subset of observed variables that were the targets of the intervention.
Generative Modeling of Molecular Dynamics Trajectories
We introduce generative modeling of molecular trajectories as a paradigm for learning flexible multi-task surrogate models of MD from data.
Revisiting Who's Harry Potter: Towards Targeted Unlearning from a Causal Intervention Perspective
This paper investigates Who's Harry Potter (WHP), a pioneering yet insufficiently understood method for LLM unlearning.
DisCo-Diff: Enhancing Continuous Diffusion Models with Discrete Latents
The discrete latents significantly simplify learning the DM's complex noise-to-data mapping by reducing the curvature of the DM's generative ODE.
Ranked #3 on
Image Generation
on ImageNet 128x128
A Recipe for Charge Density Prediction
In density functional theory, charge density is the core attribute of atomic systems from which all chemical properties can be derived.
In-Context Symmetries: Self-Supervised Learning through Contextual World Models
In this work, drawing insights from world models, we propose to instead learn a general representation that can adapt to be invariant or equivariant to different transformations by paying attention to context -- a memory module that tracks task-specific states, actions, and future states.
Verlet Flows: Exact-Likelihood Integrators for Flow-Based Generative Models
Approximations in computing model likelihoods with continuous normalizing flows (CNFs) hinder the use of these models for importance sampling of Boltzmann distributions, where exact likelihoods are required.
Deep Confident Steps to New Pockets: Strategies for Docking Generalization
Accurate blind docking has the potential to lead to new biological breakthroughs, but for this promise to be realized, docking methods must generalize well across the proteome.
Dirichlet Flow Matching with Applications to DNA Sequence Design
Further, we provide distilled Dirichlet flow matching, which enables one-step sequence generation with minimal performance hits, resulting in $O(L)$ speedups compared to autoregressive models.
AlphaFold Meets Flow Matching for Generating Protein Ensembles
When trained and evaluated on the PDB, our method provides a superior combination of precision and diversity compared to AlphaFold with MSA subsampling.
Generative Flows on Discrete State-Spaces: Enabling Multimodal Flows with Applications to Protein Co-Design
Our approach achieves state-of-the-art co-design performance while allowing the same multimodal model to be used for flexible generation of the sequence or structure.
Sample, estimate, aggregate: A recipe for causal discovery foundation models
Theoretically, we show that the model architecture is well-specified, in the sense that it can recover a causal graph consistent with graphs over subsets.
Correcting Diffusion Generation through Resampling
Despite diffusion models' superior capabilities in modeling complex distributions, there are still non-trivial distributional discrepancies between generated and ground-truth images, which has resulted in several notable problems in image generation, including missing object errors in text-to-image generation and low image quality.
Equivariant Scalar Fields for Molecular Docking with Fast Fourier Transforms
The runtime of our approach can be amortized at several levels of abstraction, and is particularly favorable for virtual screening settings with a common binding pocket.
Fast non-autoregressive inverse folding with discrete diffusion
Generating protein sequences that fold into a intended 3D structure is a fundamental step in de novo protein design.
Risk-Controlling Model Selection via Guided Bayesian Optimization
Adjustable hyperparameters of machine learning models typically impact various key trade-offs such as accuracy, fairness, robustness, or inference cost.
Removing Biases from Molecular Representations via Information Maximization
High-throughput drug screening -- using cell imaging or gene expression measurements as readouts of drug effect -- is a critical tool in biotechnology to assess and understand the relationship between the chemical structure and biological activity of a drug.
Learning Interatomic Potentials at Multiple Scales
When running MD, the MTS integrator then evaluates the smaller model for every time step and the larger model less frequently, accelerating simulation.
Particle Guidance: non-I.I.D. Diverse Sampling with Diffusion Models
In light of the widespread success of generative models, a significant amount of research has gone into speeding up their sampling time.
Harmonic Self-Conditioned Flow Matching for Multi-Ligand Docking and Binding Site Design
A significant amount of protein function requires binding small molecules, including enzymatic catalysis.
Fast protein backbone generation with SE(3) flow matching
We present FrameFlow, a method for fast protein backbone generation using SE(3) flow matching.
Compositional Sculpting of Iterative Generative Processes
A key challenge in composing iterative generative processes, such as GFlowNets and diffusion models, is that to realize the desired target distribution, all steps of the generative process need to be coordinated, and satisfy delicate balance conditions.
Artificial Intelligence for Science in Quantum, Atomistic, and Continuum Systems
Advances in artificial intelligence (AI) are fueling a new paradigm of discoveries in natural sciences.
Improving Protein Optimization with Smoothed Fitness Landscapes
The ability to engineer novel proteins with higher fitness for a desired property would be revolutionary for biotechnology and medicine.
Restart Sampling for Improving Generative Processes
Restart not only outperforms the previous best SDE results, but also accelerates the sampling speed by 10-fold / 2-fold on CIFAR-10 / ImageNet $64 \times 64$.
Towards Coherent Image Inpainting Using Denoising Diffusion Implicit Models
COPAINT also uses the Bayesian framework to jointly modify both revealed and unrevealed regions, but approximates the posterior distribution in a way that allows the errors to gradually drop to zero throughout the denoising steps, thus strongly penalizing any mismatches with the reference image.
EigenFold: Generative Protein Structure Prediction with Diffusion Models
Protein structure prediction has reached revolutionary levels of accuracy on single structures, yet distributional modeling paradigms are needed to capture the conformational ensembles and flexibility that underlie biological function.
GenPhys: From Physical Processes to Generative Models
We introduce a general family, Generative Models from Physical Processes (GenPhys), where we translate partial differential equations (PDEs) describing physical processes to generative models.
PFGM++: Unlocking the Potential of Physics-Inspired Generative Models
The new models reduce to PFGM when $D{=}1$ and to diffusion models when $D{\to}\infty$.
Ranked #1 on
Image Generation
on FFHQ 64x64 - 4x upscaling
SE(3) diffusion model with application to protein backbone generation
The design of novel protein structures remains a challenge in protein engineering for applications across biomedicine and chemistry.
Stable Target Field for Reduced Variance Score Estimation in Diffusion Models
We show that the procedure indeed helps in the challenging intermediate regime by reducing (the trace of) the covariance of training targets.
Is Conditional Generative Modeling all you need for Decision-Making?
We further demonstrate the advantages of modeling policies as conditional diffusion models by considering two other conditioning variables: constraints and skills.
Efficiently Controlling Multiple Risks with Pareto Testing
Machine learning applications frequently come with multiple diverse objectives and constraints that can change over time.
Forces are not Enough: Benchmark and Critical Evaluation for Machine Learning Force Fields with Molecular Simulations
We benchmark a collection of state-of-the-art (SOTA) ML FF models and illustrate, in particular, how the commonly benchmarked force accuracy is not well aligned with relevant simulation metrics.
DiffDock: Diffusion Steps, Twists, and Turns for Molecular Docking
We instead frame molecular docking as a generative modeling problem and develop DiffDock, a diffusion generative model over the non-Euclidean manifold of ligand poses.
Ranked #1 on
Blind Docking
on PDBbind
(Top-1 RMSD (Med.) metric)
Poisson Flow Generative Models
We interpret the data points as electrical charges on the $z=0$ hyperplane in a space augmented with an additional dimension $z$, generating a high-dimensional electric field (the gradient of the solution to Poisson equation).
Ranked #22 on
Image Generation
on CIFAR-10
Calibrated Selective Classification
Providing calibrated uncertainty estimates alongside predictions -- probabilities that correspond to true frequencies -- can be as important as having predictions that are simply accurate on average.
Antibody-Antigen Docking and Design via Hierarchical Equivariant Refinement
The binding affinity is governed by the 3D binding interface where antibody residues (paratope) closely interact with antigen residues (epitope).
Diffusion probabilistic modeling of protein backbones in 3D for the motif-scaffolding problem
Construction of a scaffold structure that supports a desired motif, conferring protein function, shows promise for the design of vaccines and enzymes.
Torsional Diffusion for Molecular Conformer Generation
Molecular conformer generation is a fundamental task in computational chemistry.
Subspace Diffusion Generative Models
Score-based models generate samples by mapping noise to data (and vice versa) via a high-dimensional diffusion process.
Ranked #17 on
Image Generation
on CIFAR-10
Simulate Time-integrated Coarse-grained Molecular Dynamics with Multi-Scale Graph Networks
Molecular dynamics (MD) simulation is essential for various scientific domains but computationally expensive.
Conformal Prediction Sets with Limited False Positives
We propose to trade coverage for a notion of precision by enforcing that the presence of incorrect candidates in the predicted conformal sets (i. e., the total number of false positives) is bounded according to a user-specified tolerance.
EquiBind: Geometric Deep Learning for Drug Binding Structure Prediction
Predicting how a drug-like molecule binds to a specific protein target is a core problem in drug discovery.
Ranked #6 on
Blind Docking
on PDBBind
Controlling Directions Orthogonal to a Classifier
We propose to identify directions invariant to a given classifier so that these directions can be controlled in tasks such as style transfer.
Independent SE(3)-Equivariant Models for End-to-End Rigid Protein Docking
Protein complex formation is a central problem in biology, being involved in most of the cell's processes, and essential for applications, e. g. drug design or protein engineering.
Fragment-based Sequential Translation for Molecular Optimization
Equipped with the learned fragment vocabulary, we propose Fragment-based Sequential Translation (FaST), which learns a reinforcement learning (RL) policy to iteratively translate model-discovered molecules into increasingly novel molecules while satisfying desired properties.
Learning Representations that Support Robust Transfer of Predictors
We further demonstrate the impact of optimizing such transfer risk on two controlled settings, each representing a different pattern of environment shift, as well as on two real-world datasets.
Crystal Diffusion Variational Autoencoder for Periodic Material Generation
Generating the periodic structure of stable materials is a long-standing challenge for the material design community.
Iterative Refinement Graph Neural Network for Antibody Sequence-Structure Co-design
In this paper, we propose a generative model to automatically design the CDRs of antibodies with enhanced binding specificity or neutralization capabilities.
Learning Task Informed Abstractions
Current model-based reinforcement learning methods struggle when operating from complex visual scenes due to their inability to prioritize task-relevant features.
Model-based Reinforcement Learning
reinforcement-learning
+1
Mol2Image: Improved Conditional Flow Models for Molecule to Image Synthesis
In this paper, we aim to synthesize cell microscopy images under different molecular interventions, motivated by practical applications to drug development.
Consistent Accelerated Inference via Confident Adaptive Transformers
In this work, we present CATs -- Confident Adaptive Transformers -- in which we simultaneously increase computational efficiency, while guaranteeing a specifiable degree of consistency with the original model with high confidence.
Few-shot Conformal Prediction with Auxiliary Tasks
We develop a novel approach to conformal prediction when the target task has limited data available for training.
Discovering Synergistic Drug Combinations for COVID with Biological Bottleneck Models
Drug combinations play an important role in therapeutics due to its better efficacy and reduced toxicity.
Information Obfuscation of Graph Neural Networks
While the advent of Graph Neural Networks (GNNs) has greatly improved node and graph representation learning in many applications, the neighborhood aggregation scheme exposes additional vulnerabilities to adversaries seeking to extract node-level information about sensitive attributes.
Efficient Conformal Prediction via Cascaded Inference with Expanded Admission
This set is guaranteed to contain a correct answer with high probability, and is well-suited for many open-ended classification tasks.
Improved Conditional Flow Models for Molecule to Image Synthesis
In this paper, we aim to synthesize cell microscopy images under different molecular interventions, motivated by practical applications to drug development.
Optimal Transport Graph Neural Networks
Current graph neural network (GNN) architectures naively average or sum node embeddings into an aggregated graph representation -- potentially losing structural or semantic information.
Ranked #1 on
Drug Discovery
on BBBP
Enforcing Predictive Invariance across Structured Biomedical Domains
We evaluate our method on multiple applications: molecular property prediction, protein homology and stability prediction and show that RGM significantly outperforms previous state-of-the-art baselines.
Adaptive Invariance for Molecule Property Prediction
Effective property prediction methods can help accelerate the search for COVID-19 antivirals either through accurate in-silico screens or by effectively guiding on-going at-scale experimental efforts.
Self-Supervised Learning of Appliance Usage
We show that this cross-modal prediction task allows us to detect when a particular appliance is used, and the location of the appliance in the home, all in a self-supervised manner, without any labeled data.
The Benefits of Pairwise Discriminators for Adversarial Training
We provide sufficient conditions for local convergence; characterize the capacity balance that should guide the discriminator and generator choices; and construct examples of minimally sufficient discriminators.
Generalization and Representational Limits of Graph Neural Networks
We address two fundamental questions about graph neural networks (GNNs).
Improving Molecular Design by Stochastic Iterative Target Augmentation
The property predictor is then used as a likelihood model for filtering candidate structures from the generative model.
Multi-Objective Molecule Generation using Interpretable Substructures
These rationales are identified from molecules as substructures that are likely responsible for each property of interest.
Blank Language Models
We propose Blank Language Model (BLM), a model that generates sequences by dynamically creating and filling in blanks.
Hierarchical Generation of Molecular Graphs using Structural Motifs
Indeed, as we demonstrate, their performance degrades significantly for larger molecules.
Generative Models for Graph-Based Protein Design
Engineered proteins offer the potential to solve many problems in biomedicine, energy, and materials science, but creating designs that succeed is difficult in practice.
Direct Optimization through \arg \max for Discrete Variational Auto-Encoder
Reparameterization of variational auto-encoders with continuous random variables is an effective method for reducing the variance of their gradient estimates.
Iterative Target Augmentation for Effective Conditional Generation
Many challenging prediction problems, from molecular optimization to program synthesis, involve creating complex structured objects as outputs.
Hierarchical Graph-to-Graph Translation for Molecules
The problem of accelerating drug discovery relies heavily on automatic tools to optimize precursor molecules to afford them with better biochemical properties.
Ranked #1 on
Drug Discovery
on QED
Solving graph compression via optimal transport
The transport problem is seeded with prior information about node importance, attributes, and edges in the graph.
Path-Augmented Graph Transformer Network
Much of the recent work on learning molecular representations has been based on Graph Convolution Networks (GCN).
Strategic Prediction with Latent Aggregative Games
We introduce a new class of context dependent, incomplete information games to serve as structured prediction models for settings with significant strategic interactions.
Alignment Based Mathching Networks for One-Shot Classification and Open-Set Recognition
An attention mechanism can be used to highlight the area of the image that the model focuses on thus offering a narrow view into the mechanism of classification.
Learning Multimodal Graph-to-Graph Translation for Molecule Optimization
We evaluate our model on multiple molecule optimization tasks and show that our model outperforms previous state-of-the-art baselines by a significant margin.
Analyzing Learned Molecular Representations for Property Prediction
In addition, we introduce a graph convolutional model that consistently matches or outperforms models using fixed molecular descriptors as well as previous graph neural architectures on both public and proprietary datasets.
Ranked #3 on
Molecular Property Prediction
on QM9
Alignment Based Matching Networks for One-Shot Classification and Open-Set Recognition
An attention mechanism can be used to highlight the area of the image that the model focuses on thus offering a narrow view into the mechanism of classification.
Learning Multimodal Graph-to-Graph Translation for Molecular Optimization
We evaluate our model on multiple molecular optimization tasks and show that our model outperforms previous state-of-the-art baselines.
Towards Robust Interpretability with Self-Explaining Neural Networks
Most recent work on interpretability of complex machine learning models has focused on estimating a-posteriori explanations for previously trained models around specific predictions.
The Variational Homoencoder: Learning to learn high capacity generative models from few examples
However, when this generative model is expressed as a powerful neural network such as a PixelCNN, we show that existing learning techniques typically fail to effectively use latent variables.
Direct Optimization through $\arg \max$ for Discrete Variational Auto-Encoder
We demonstrate empirically the effectiveness of the direct loss minimization technique in variational autoencoders with both unstructured and structured discrete latent variables.
Junction Tree Variational Autoencoder for Molecular Graph Generation
We evaluate our model on multiple tasks ranging from molecular generation to optimization.
Ranked #1 on
Molecular Graph Generation
on InterBioScreen
The Variational Homoencoder: Learning to Infer High-Capacity Generative Models from Few Examples
Hierarchical Bayesian methods have the potential to unify many related tasks (e. g. k-shot classification, conditional, and unconditional generation) by framing each as inference within a single generative model.
Using Deep Reinforcement Learning to Generate Rationales for Molecules
Deep learning algorithms are increasingly used in modeling chemical processes.
Local Aggregative Games
Aggregative games provide a rich abstraction to model strategic multi-agent interactions.
Predicting Organic Reaction Outcomes with Weisfeiler-Lehman Network
The prediction of organic reaction outcomes is a fundamental problem in computational chemistry.
Sequence to Better Sequence: Continuous Revision of Combinatorial Structures
Under this model, gradient methods can be used to efficiently optimize the continuous latent factors with respect to inferred outcomes.
Grounding Language for Transfer in Deep Reinforcement Learning
In this paper, we explore the utilization of natural language to drive transfer for reinforcement learning (RL).
Style Transfer from Non-Parallel Text by Cross-Alignment
We demonstrate the effectiveness of this cross-alignment method on three tasks: sentiment modification, decipherment of word substitution ciphers, and recovery of word order.
Ranked #7 on
Text Style Transfer
on Yelp Review Dataset (Small)
Deriving Neural Architectures from Sequence and Graph Kernels
The design of neural architectures for structured objects is typically guided by experimental insights rather than a formal process.
Aspect-augmented Adversarial Networks for Domain Adaptation
We introduce a neural method for transfer learning between two (source and target) classification tasks or aspects over the same domain.
Learning Tree Structured Potential Games
Many real phenomena, including behaviors, involve strategic interactions that can be learned from data.
Learning Optimal Interventions
Assuming the underlying relationship remains invariant under intervention, we develop efficient algorithms to identify the optimal intervention policy from limited data and provide theoretical guarantees for our approach in a Gaussian Process setting.
Rationalizing Neural Predictions
Our approach combines two modular components, generator and encoder, which are trained to operate well together.
High Dimensional Inference with Random Maximum A-Posteriori Perturbations
This paper shows that the expected value of perturb-max inference with low dimensional perturbations can be used sequentially to generate unbiased samples from the Gibbs distribution.
Semi-supervised Question Retrieval with Gated Convolutions
Question answering forums are rapidly growing in size with no effective automated ability to refer to and reuse answers already available for previous posted questions.
Principal Differences Analysis: Interpretable Characterization of Differences between Distributions
We introduce principal differences analysis (PDA) for analyzing differences between high-dimensional distributions.
Steps Toward Deep Kernel Methods from Infinite Neural Networks
Contemporary deep neural networks exhibit impressive results on practical problems.
Molding CNNs for text: non-linear, non-consecutive convolutions
Moreover, we extend the n-gram convolution to non-consecutive words to recognize patterns with intervening words.
Structured Prediction: From Gaussian Perturbations to Linear-Time Principled Algorithms
Thus, using the maximum loss over random structured outputs is a principled way of learning the parameter of structured prediction models.
CRAFT: ClusteR-specific Assorted Feature selecTion
We present a framework for clustering with cluster-specific feature selection.
An Unsupervised Method for Uncovering Morphological Chains
In contrast, we propose a model for unsupervised morphological analysis that integrates orthographic and semantic views of words.
Controlling privacy in recommender systems
Recommender systems involve an inherent trade-off between accuracy of recommendations and the extent to which users are willing to release information about their preferences.
Learning Efficient Random Maximum A-Posteriori Predictors with Non-Decomposable Loss Functions
In this work we develop efficient methods for learning random MAP predictors for structured label problems.
On Measure Concentration of Random Maximum A-Posteriori Perturbations
Applying the general result to MAP perturbations can yield a more efficient algorithm to approximate sampling from the Gibbs distribution.
On Sampling from the Gibbs Distribution with Random Maximum A-Posteriori Perturbations
In this paper we describe how MAP inference can be used to sample efficiently from Gibbs distributions.
Proceedings of the Twenty-First Conference on Uncertainty in Artificial Intelligence (2005)
This is the Proceedings of the Twenty-First Conference on Uncertainty in Artificial Intelligence, which was held in Edinburgh, Scotland July 26 - 29 2005.
On the Statistical Efficiency of $\ell_{1,p}$ Multi-Task Learning of Gaussian Graphical Models
In this paper, we present $\ell_{1, p}$ multi-task structure learning for Gaussian graphical models.
