Kevin Murphy 0002

ICML Conference 2025 Conference Paper

Direct Motion Models for Assessing Generated Videos

• Kelsey R. Allen
• Carl Doersch
• Guangyao Zhou
• Mohammed Suhail
• Danny Driess
• Ignacio Rocco
• Yulia Rubanova
• Thomas Kipf

A current limitation of video generative video models is that they generate plausible looking frames, but poor motion — an issue that is not well captured by FVD and other popular methods for evaluating generated videos. Here we go beyond FVD by developing a metric which better measures plausible object interactions and motion. Our novel approach is based on auto-encoding point tracks and yields motion features that can be used to not only compare distributions of videos (as few as one generated and one ground truth, or as many as two datasets), but also for evaluating motion of single videos. We show that using point tracks instead of pixel reconstruction or action recognition features results in a metric which is markedly more sensitive to temporal distortions in synthetic data, and can predict human evaluations of temporal consistency and realism in generated videos obtained from open-source models better than a wide range of alternatives. We also show that by using a point track representation, we can spatiotemporally localize generative video inconsistencies, providing extra interpretability of generated video errors relative to prior work. An overview of the results and link to the code can be found on the project page: trajan-paper. github. io.

ICML Conference 2025 Conference Paper

Distributional Diffusion Models with Scoring Rules

• Valentin De Bortoli
• Alexandre Galashov
• J. Swaroop Guntupalli
• Guangyao Zhou
• Kevin Murphy 0002
• Arthur Gretton
• Arnaud Doucet

Diffusion models generate high-quality synthetic data. They operate by defining a continuous-time forward process which gradually adds Gaussian noise to data until fully corrupted. The corresponding reverse process progressively “denoises" a Gaussian sample into a sample from the data distribution. However, generating high-quality outputs requires many discretization steps to obtain a faithful approximation of the reverse process. This is expensive and has motivated the development of many acceleration methods. We propose to speed up sample generation by learning the posterior distribution of clean data samples given their noisy versions, instead of only the mean of this distribution. This allows us to sample from the probability transitions of the reverse process on a coarse time scale, significantly accelerating inference with minimal degradation of the quality of the output. This is accomplished by replacing the standard regression loss used to estimate conditional means with a scoring rule. We validate our method on image and robot trajectory generation, where we consistently outperform standard diffusion models at few discretization steps.

ICML Conference 2025 Conference Paper

Improving Transformer World Models for Data-Efficient RL

• Antoine Dedieu
• Joseph Ortiz
• Xinghua Lou
• Carter Wendelken
• J. Swaroop Guntupalli
• Wolfgang Lehrach
• Miguel Lázaro-Gredilla
• Kevin Murphy 0002

We present an approach to model-based RL that achieves a new state of the art performance on the challenging Craftax-classic benchmark, an open-world 2D survival game that requires agents to exhibit a wide range of general abilities—such as strong generalization, deep exploration, and long-term reasoning. With a series of careful design choices aimed at improving sample efficiency, our MBRL algorithm achieves a reward of 69. 66% after only 1M environment steps, significantly outperforming DreamerV3, which achieves $53. 2%$, and, for the first time, exceeds human performance of 65. 0%. Our method starts by constructing a SOTA model-free baseline, using a novel policy architecture that combines CNNs and RNNs. We then add three improvements to the standard MBRL setup: (a) "Dyna with warmup", which trains the policy on real and imaginary data, (b) "nearest neighbor tokenizer" on image patches, which improves the scheme to create the transformer world model (TWM) inputs, and (c) "block teacher forcing", which allows the TWM to reason jointly about the future tokens of the next timestep.

ICML Conference 2025 Conference Paper

Towards a Mechanistic Explanation of Diffusion Model Generalization

• Matthew Niedoba
• Berend Zwartsenberg
• Kevin Murphy 0002
• Frank Wood

We propose a simple, training-free mechanism which explains the generalization behaviour of diffusion models. By comparing pre-trained diffusion models to their theoretically optimal empirical counterparts, we identify a shared local inductive bias across a variety of network architectures. From this observation, we hypothesize that network denoisers generalize through localized denoising operations, as these operations approximate the training objective well over much of the training distribution. To validate our hypothesis, we introduce novel denoising algorithms which aggregate local empirical denoisers to replicate network behaviour. Comparing these algorithms to network denoisers across forward and reverse diffusion processes, our approach exhibits consistent visual similarity to neural network outputs, with lower mean squared error than previously proposed methods.

ICML Conference 2024 Conference Paper

Outlier-robust Kalman Filtering through Generalised Bayes

• Gerardo Duran-Martin
• Matías Altamirano
• Alexander Y. Shestopaloff
• Leandro Sánchez-Betancourt
• Jeremias Knoblauch
• Matt Jones 0002
• François-Xavier Briol
• Kevin Murphy 0002

We derive a novel, provably robust, efficient, and closed-form Bayesian update rule for online filtering in state-space models in the presence of outliers and misspecified measurement models. Our method combines generalised Bayesian inference with filtering methods such as the extended and ensemble Kalman filter. We use the former to show robustness and the latter to ensure computational efficiency in the case of nonlinear models. Our method matches or outperforms other robust filtering methods (such as those based on variational Bayes) at a much lower computational cost. We show this empirically on a range of filtering problems with outlier measurements, such as object tracking, state estimation in high-dimensional chaotic systems, and online learning of neural networks.

ICML Conference 2023 Conference Paper

Muse: Text-To-Image Generation via Masked Generative Transformers

• Huiwen Chang
• Han Zhang 0010
• Jarred Barber
• Aaron Maschinot
• José Lezama
• Lu Jiang 0004
• Ming-Hsuan Yang 0001
• Kevin Murphy 0002

We present Muse, a text-to-image Transformermodel that achieves state-of-the-art image genera-tion performance while being significantly moreefficient than diffusion or autoregressive models. Muse is trained on a masked modeling task indiscrete token space: given the text embeddingextracted from a pre-trained large language model(LLM), Muse learns to predict randomly maskedimage tokens. Compared to pixel-space diffusionmodels, such as Imagen and DALL-E 2, Muse issignificantly more efficient due to the use of dis-crete tokens and requires fewer sampling itera-tions; compared to autoregressive models such asParti, Muse is more efficient due to the use of par-allel decoding. The use of a pre-trained LLM en-ables fine-grained language understanding, whichtranslates to high-fidelity image generation andthe understanding of visual concepts such as ob-jects, their spatial relationships, pose, cardinalityetc. Our 900M parameter model achieves a newSOTA on CC3M, with an FID score of 6. 06. TheMuse 3B parameter model achieves an FID of7. 88 on zero-shot COCO evaluation, along with aCLIP score of 0. 32. Muse also directly enables anumber of image editing applications without theneed to fine-tune or invert the model: inpainting, outpainting, and mask-free editing. More resultsand videos demonstrating editing are available at https: //muse-icml. github. io/

UAI Conference 2020 Conference Paper

Amortized Bayesian Optimization over Discrete Spaces

• Kevin Swersky
• Yulia Rubanova
• David Dohan
• Kevin Murphy 0002

Bayesian optimization is a principled approach for globally optimizing expensive, black-box functions by using a surrogate model of the objective. However, each step of Bayesian optimization involves solving an inner optimization problem, in which we maximize an acquisition function derived from the surrogate model to decide where to query next. This inner problem can be challenging to solve, particularly in discrete spaces, such as protein sequences or molecular graphs, where gradient-based optimization cannot be used. Our key insight is that we can train a generative model to generate candidates that maximize the acquisition function. This is faster than standard model-free local search methods, since we can amortize the cost of learning the model across multiple rounds of Bayesian optimization. We therefore call this Amortized Bayesian Optimization. On several challenging discrete design problems, we show this method generally outperforms other methods at optimizing the inner acquisition function, resulting in more efficient optimization of the outer black-box objective.

ICML Conference 2020 Conference Paper

Collapsed Amortized Variational Inference for Switching Nonlinear Dynamical Systems

• Zhe Dong
• Bryan A. Seybold
• Kevin Murphy 0002
• Hung H. Bui

We propose an efficient inference method for switching nonlinear dynamical systems. The key idea is to learn an inference network which can be used as a proposal distribution for the continuous latent variables, while performing exact marginalization of the discrete latent variables. This allows us to use the reparameterization trick, and apply end-to-end training with stochastic gradient descent. We show that the proposed method can successfully segment time series data, including videos and 3D human pose, into meaningful “regimes” by using the piece-wise nonlinear dynamics.

ICLR Conference 2020 Conference Paper

Model-based reinforcement learning for biological sequence design

• Christof Angermüller
• David Dohan
• David Belanger 0002
• Ramya Deshpande
• Kevin Murphy 0002
• Lucy J. Colwell

The ability to design biological structures such as DNA or proteins would have considerable medical and industrial impact. Doing so presents a challenging black-box optimization problem characterized by the large-batch, low round setting due to the need for labor-intensive wet lab evaluations. In response, we propose using reinforcement learning (RL) based on proximal-policy optimization (PPO) for biological sequence design. RL provides a flexible framework for optimization generative sequence models to achieve specific criteria, such as diversity among the high-quality sequences discovered. We propose a model-based variant of PPO, DyNA-PPO, to improve sample efficiency, where the policy for a new round is trained offline using a simulator fit on functional measurements from prior rounds. To accommodate the growing number of observations across rounds, the simulator model is automatically selected at each round from a pool of diverse models of varying capacity. On the tasks of designing DNA transcription factor binding sites, designing antimicrobial proteins, and optimizing the energy of Ising models based on protein structure, we find that DyNA-PPO performs significantly better than existing methods in settings in which modeling is feasible, while still not performing worse in situations in which a reliable model cannot be learned.

ICML Conference 2020 Conference Paper

Population-Based Black-Box Optimization for Biological Sequence Design

• Christof Angermüller
• David Belanger 0002
• Andreea Gane
• Zelda Mariet
• David Dohan
• Kevin Murphy 0002
• Lucy J. Colwell
• D. Sculley

The use of black-box optimization for the design of new biological sequences is an emerging research area with potentially revolutionary impact. The cost and latency of wet-lab experiments requires methods that find good sequences in few experimental rounds of large batches of sequences — a setting that off-the-shelf black-box optimization methods are ill-equipped to handle. We find that the performance of existing methods varies drastically across optimization tasks, posing a significant obstacle to real-world applications. To improve robustness, we propose Population-Based Black-Box Optimization (P3BO), which generates batches of sequences by sampling from an ensemble of methods. The number of sequences sampled from any method is proportional to the quality of sequences it previously proposed, allowing P3BO to combine the strengths of individual methods while hedging against their innate brittleness. Adapting the hyper-parameters of each of the methods online using evolutionary optimization further improves performance. Through extensive experiments on in-silico optimization tasks, we show that P3BO outperforms any single method in its population, proposing higher quality sequences as well as more diverse batches. As such, P3BO and Adaptive-P3BO are a crucial step towards deploying ML to real-world sequence design.

ICML Conference 2019 Conference Paper

NAS-Bench-101: Towards Reproducible Neural Architecture Search

• Chris Ying
• Aaron Klein
• Eric Christiansen
• Esteban Real
• Kevin Murphy 0002
• Frank Hutter

Recent advances in neural architecture search (NAS) demand tremendous computational resources, which makes it difficult to reproduce experiments and imposes a barrier-to-entry to researchers without access to large-scale computation. We aim to ameliorate these problems by introducing NAS-Bench-101, the first public architecture dataset for NAS research. To build NAS-Bench-101, we carefully constructed a compact, yet expressive, search space, exploiting graph isomorphisms to identify 423k unique convolutional architectures. We trained and evaluated all of these architectures multiple times on CIFAR-10 and compiled the results into a large dataset of over 5 million trained models. This allows researchers to evaluate the quality of a diverse range of models in milliseconds by querying the pre-computed dataset. We demonstrate its utility by analyzing the dataset as a whole and by benchmarking a range of architecture optimization algorithms.

ICML Conference 2018 Conference Paper

Fixing a Broken ELBO

• Alexander A. Alemi
• Ben Poole
• Ian Fischer
• Joshua V. Dillon
• Rif A. Saurous
• Kevin Murphy 0002

Recent work in unsupervised representation learning has focused on learning deep directed latentvariable models. Fitting these models by maximizing the marginal likelihood or evidence is typically intractable, thus a common approximation is to maximize the evidence lower bound (ELBO) instead. However, maximum likelihood training (whether exact or approximate) does not necessarily result in a good latent representation, as we demonstrate both theoretically and empirically. In particular, we derive variational lower and upper bounds on the mutual information between the input and the latent variable, and use these bounds to derive a rate-distortion curve that characterizes the tradeoff between compression and reconstruction accuracy. Using this framework, we demonstrate that there is a family of models with identical ELBO, but different quantitative and qualitative characteristics. Our framework also suggests a simple new method to ensure that latent variable models with powerful stochastic decoders do not ignore their latent code.

ICML Conference 2011 Conference Paper

Piecewise Bounds for Estimating Bernoulli-Logistic Latent Gaussian Models

• Benjamin M. Marlin
• Mohammad Emtiyaz Khan
• Kevin Murphy 0002

UAI Conference 2009 Conference Paper

Group Sparse Priors for Covariance Estimation

• Benjamin M. Marlin
• Mark Schmidt 0001
• Kevin Murphy 0002

Recently it has become popular to learn sparse Gaussian graphical models (GGMs) by imposing ℓ1 or group ℓ1, 2 penalties on the elements of the precision matrix. This penalized likelihood approach results in a tractable convex optimization problem. In this paper, we reinterpret these results as performing MAP estimation under a novel prior which we call the group ℓ1 and ℓ1, 2 positivedefinite matrix distributions. This enables us to build a hierarchical model in which the ℓ1 regularization terms vary depending on which group the entries are assigned to, which in turn allows us to learn block structured sparse GGMs with unknown group assignments. Exact inference in this hierarchical model is intractable, due to the need to compute the normalization constant of these matrix distributions. However, we derive upper bounds on the partition functions, which lets us use fast variational inference (optimizing a lower bound on the joint posterior). We show that on two real world data sets (motion capture and financial data), our method which infers the block structure outperforms a method that uses a fixed block structure, which in turn outperforms baseline methods that ignore block structure.

UAI Conference 2009 Conference Paper

Modeling Discrete Interventional Data using Directed Cyclic Graphical Models

• Mark Schmidt 0001
• Kevin Murphy 0002

We outline a representation for discrete multivariate distributions in terms of interventional potential functions that are globally normalized. This representation can be used to model the effects of interventions, and the independence properties encoded in this model can be represented as a directed graph that allows cycles. In addition to discussing inference and sampling with this representation, we give an exponential family parametrization that allows parameter estimation to be stated as a convex optimization problem; we also give a convex relaxation of the task of simultaneous parameter and structure learning using group `1 regularization. The model is evaluated on simulated data and intracellular flow cytometry data.

ICML Conference 2009 Conference Paper

Sparse Gaussian graphical models with unknown block structure

• Benjamin M. Marlin
• Kevin Murphy 0002

UAI Conference 2007 Conference Paper

Bayesian structure learning using dynamic programming and MCMC

• Daniel Eaton
• Kevin Murphy 0002

MCMC methods for sampling from the space of DAGs can mix poorly due to the local nature of the proposals that are commonly used. It has been shown that sampling from the space of node orders yields better results [FK03, EW06]. Recently, Koivisto and Sood showed how one can analytically marginalize over orders using dynamic programming (DP) [KS04, Koi06]. Their method computes the exact marginal posterior edge probabilities, thus avoiding the need for MCMC. Unfortunately, there are four drawbacks to the DP technique: it can only use modular priors, it can only compute posteriors over modular features, it is difficult to compute a predictive density, and it takes exponential time and space. We show how to overcome the first three of these problems by using the DP algorithm as a proposal distribution for MCMC in DAG space. We show that this hybrid technique converges to the posterior faster than other methods, resulting in more accurate structure learning and higher predictive likelihoods on test data.

ICML Conference 2007 Conference Paper

Modeling changing dependency structure in multivariate time series

• Xiang Xuan
• Kevin Murphy 0002

We show how to apply the efficient Bayesian changepoint detection techniques of Fearnhead in the multivariate setting. We model the joint density of vector-valued observations using undirected Gaussian graphical models, whose structure we estimate. We show how we can exactly compute the MAP segmentation, as well as how to draw perfect samples from the posterior over segmentations, simultaneously accounting for uncertainty about the number and location of changepoints, as well as uncertainty about the covariance structure. We illustrate the technique by applying it to financial data and to bee tracking data.

ICML Conference 2006 Conference Paper

Accelerated training of conditional random fields with stochastic gradient methods

• S. V. N. Vishwanathan
• Nicol N. Schraudolph
• Mark Schmidt 0001
• Kevin Murphy 0002

We apply Stochastic Meta-Descent (SMD), a stochastic gradient optimization method with gain vector adaptation, to the training of Conditional Random Fields (CRFs). On several large data sets, the resulting optimizer converges to the same quality of solution over an order of magnitude faster than limited-memory BFGS, the leading method reported to date. We report results for both exact and inexact inference techniques.

ICRA Conference 2004 Conference Paper

Representing Hierarchical POMDPs as DBNs for Multi-scale Robot Localization

• Georgios Theocharous
• Kevin Murphy 0002
• Leslie Pack Kaelbling

We explore the advantages of representing hierarchical partially observable Markov decision processes (H-POMDPs) as dynamic Bayesian networks (DBNs). In particular, we focus on the special case of using H-POMDPs to represent multi-resolution spatial maps for indoor robot navigation. Our results show that a DBN representation of H-POMDPs can train significantly faster than the original learning algorithm for H-POMDPs or the equivalent flat POMDP, and requires much less data. In addition, the DBN formulation can easily be extended to parameter tying and factoring of variables, which further reduces the time and sample complexity. This enables us to apply H-POMDP methods to much larger problems than previously possible.

UAI Conference 2001 Conference Paper

The Factored Frontier Algorithm for Approximate Inference in DBNs

• Kevin Murphy 0002
• Yair Weiss

The Factored Frontier (FF) algorithm is a simple approximate inferencealgorithm for Dynamic Bayesian Networks (DBNs). It is very similar tothe fully factorized version of the Boyen-Koller (BK) algorithm, butinstead of doing an exact update at every step followed bymarginalisation (projection), it always works with factoreddistributions. Hence it can be applied to models for which the exactupdate step is intractable. We show that FF is equivalent to (oneiteration of) loopy belief propagation (LBP) on the original DBN, andthat BK is equivalent (to one iteration of) LBP on a DBN where wecluster some of the nodes. We then show empirically that byiterating, LBP can improve on the accuracy of both FF and BK. Wecompare these algorithms on two real-world DBNs: the first is a modelof a water treatment plant, and the second is a coupled HMM, used tomodel freeway traffic.

UAI Conference 2000 Conference Paper

Rao-Blackwellised Particle Filtering for Dynamic Bayesian Networks

• Arnaud Doucet
• Nando de Freitas
• Kevin Murphy 0002
• Stuart Russell 0001

Particle filters (PFs) are powerful sampling-based inference/learning algorithms for dynamic Bayesian networks (DBNs). They allow us to treat, in a principled way, any type of probability distribution, nonlinearity and non-stationarity. They have appeared in several fields under such names as ``condensation'', ``sequential Monte Carlo'' and ``survival of the fittest''. In this paper, we show how we can exploit the structure of the DBN to increase the efficiency of particle filtering, using a technique known as Rao-Blackwellisation. Essentially, this samples some of the variables, and marginalizes out the rest exactly, using the Kalman filter, HMM filter, junction tree algorithm, or any other finite dimensional optimal filter. We show that Rao-Blackwellised particle filters (RBPFs) lead to more accurate estimates than standard PFs. We demonstrate RBPFs on two problems, namely non-stationary online regression with radial basis function networks and robot localization and map building. We also discuss other potential application areas and provide references to some finite dimensional optimal filters.

UAI Conference 1999 Conference Paper

A Variational Approximation for Bayesian Networks with Discrete and Continuous Latent Variables

• Kevin Murphy 0002

We show how to use a variational approximation to the logistic function to perform approximate inference in Bayesian networks containing discrete nodes with continuous parents. Essentially, we convert the logistic function to a Gaussian, which facilitates exact inference, and then iteratively adjust the variational parameters to improve the quality of the approximation. We demonstrate experimentally that this approximation is faster and potentially more accurate than sampling. We also introduce a simple new technique for handling evidence, which allows us to handle arbitrary distributions on observed nodes, as well as achieving a significant speedup in networks with discrete variables of large cardinality.

UAI Conference 1999 Conference Paper

Loopy Belief Propagation for Approximate Inference: An Empirical Study

• Kevin Murphy 0002
• Yair Weiss
• Michael I. Jordan

Recently, researchers have demonstrated that loopy belief propagation --- the use of Pearls polytree algorithm IN a Bayesian network WITH loops OF error- correcting codes.The most dramatic instance OF this IS the near Shannon - limit performance OF Turbo Codes codes whose decoding algorithm IS equivalent TO loopy belief propagation IN a chain - structured Bayesian network. IN this paper we ask : IS there something special about the error - correcting code context, OR does loopy propagation WORK AS an approximate inference schemeIN a more general setting? We compare the marginals computed using loopy propagation TO the exact ones IN four Bayesian network architectures, including two real - world networks : ALARM AND QMR.We find that the loopy beliefs often converge AND WHEN they do, they give a good approximation TO the correct marginals.However,ON the QMR network, the loopy beliefs oscillated AND had no obvious relationship TO the correct posteriors. We present SOME initial investigations INTO the cause OF these oscillations, AND show that SOME simple methods OF preventing them lead TO the wrong results.

UAI Conference 1998 Conference Paper

Learning the Structure of Dynamic Probabilistic Networks

• Nir Friedman
• Kevin Murphy 0002
• Stuart Russell 0001

Dynamic probabilistic networks are a compact representation of complex stochastic processes. In this paper we examine how to learn the structure of a DPN from data. We extend structure scoring rules for standard probabilistic networks to the dynamic case, and show how to search for structure when some of the variables are hidden. Finally, we examine two applications where such a technology might be useful: predicting and classifying dynamic behaviors, and learning causal orderings in biological processes. We provide empirical results that demonstrate the applicability of our methods in both domains.