Jeff Schneider

AAMAS Conference 2026 Conference Paper

ME-IGM: Individual-Global-Max in Maximum Entropy Multi-Agent Reinforcement Learning

• Wentse Chen
• Yuxuan Li
• Shiyu Huang
• Jiayu Chen
• Jeff Schneider

Multi-agent credit assignment is a fundamental challenge for cooperative multi-agent reinforcement learning (MARL), where a team of agents learn from shared reward signals. The Individual-Global- Max (IGM) condition is a widely used principle for multi-agent credit assignment, requiring that the joint action determined by individual Q-functions maximizes the global Q-value. Meanwhile, the principle of maximum entropy has been leveraged to enhance exploration in MARL. However, we identify a critical limitation in existing maximum entropy MARL methods: a misalignment arises between local policies and the joint policy that maximizes the global Q-value, leading to violations of the IGM condition. To address this misalignment, we propose an order-preserving transformation. Building on it, we introduce ME-IGM, a novel maximum entropy MARL algorithm compatible with any credit assignment mechanism that satisfies the IGM condition while enjoying the benefits of maximum entropy exploration. We empirically evaluate two variants of ME-IGM: ME-QMIX and ME-QPLEX, in non-monotonic matrix games, and demonstrate their state-of-the-art performance across 17 scenarios in SMAC-v2 and Overcooked.

TMLR Journal 2025 Journal Article

Beyond Parameter Count: Implicit Bias in Soft Mixture of Experts

• Youngseog Chung
• Dhruv Malik
• Jeff Schneider
• Yuanzhi Li
• Aarti Singh

The traditional viewpoint on Sparse Mixture of Experts (MoE) models is that instead of training a single _large_ expert, which is computationally expensive, we can train many _small_ experts. The hope is that if the total parameter count of the small experts equals that of the singular large expert, then we retain the representation power of the large expert while gaining computational tractability and promoting expert specialization. The recently introduced Soft MoE replaces the Sparse MoE's discrete routing mechanism with a differentiable gating function that smoothly mixes tokens. While this smooth gating function successfully mitigates the various training instabilities associated with Sparse MoE, it is unclear whether it induces implicit biases that affect Soft MoE's representation power or potential for expert specialization. We prove that Soft MoE with a single arbitrarily powerful expert cannot represent simple convex functions. This justifies that Soft MoE's success cannot be explained by the traditional viewpoint of many small experts collectively mimicking the representation power of a single large expert, and that multiple experts are actually _necessary_ to achieve good representation power (even for a fixed total parameter count). Continuing along this line of investigation, we introduce a notion of expert specialization for Soft MoE, and while varying the number of experts yet fixing the total parameter count, we consider the following (computationally intractable) task. Given any input, how can we discover the expert subset that is specialized to predict this input's label? We empirically show that when there are many small experts, the architecture is implicitly biased in a fashion that allows us to efficiently approximate the specialized expert subset. Our method can be easily implemented to potentially reduce computation during inference. For example, using our method on ImageNet, one can perform inference using only $1/8$ of the experts and still retain $99$% of the test accuracy of using all experts.

NeurIPS Conference 2025 Conference Paper

Efficient Bayesian Experiment Design with Equivariant Networks

• Conor Igoe
• Tejus Gupta
• Jeff Schneider

Recent work in Bayesian Experiment Design (BED) has shown the value of using Deep Learning (DL) to obtain highly efficient adaptive experiment designs. In this paper, we argue that a central bottleneck of DL training for BED is belief explosion. Specifically, as an agent progresses deeper into an experiment, the effective number of realisable beliefs grows enormously, placing significant sampling burdens on offline training schemes in an effort to gather experience from all regions of belief space. We argue that choosing an appropriate inductive bias for actor/critic networks is a critical component in mitigating the effects of belief explosion and has so far been overlooked in the BED literature. We show how Graph Neural Networks are particularly well-suited for BED DL training due to their domain permutation equivariance properties, resulting in multiple orders of magnitude improvement to sample efficiency compared to naive parameterizations.

NeurIPS Conference 2025 Conference Paper

Improving Model-Based Reinforcement Learning by Converging to Flatter Minima

• Shrinivas Ramasubramanian
• Benjamin Freed
• Alexandre Capone
• Jeff Schneider

Model-based reinforcement learning (MBRL) hinges on a learned dynamics model whose errors can compound along imagined rollouts. We study how encouraging \emph{flatness} in the model’s training loss affects downstream control, and show that steering optimization toward flatter minima yields a better policy. Concretely, we integrate \emph{Sharpness-Aware Minimization} (SAM) into world-model training as a drop-in objective, leaving the planner and policy components unchanged. On the theory side, we derive PAC-Bayesian bounds that link first-order sharpness to the value-estimation gap and the performance gap between model-optimal and true-optimal policies, implying that flatter minima tighten both. Empirically, SAM reduces measured sharpness and value-prediction error and improves returns across HumanoidBench, Atari-100k, and high-DoF DeepMind Control tasks. Augmenting existing MBRL algorithms with SAM increases mean return, with especially large gains in settings with high dimensional state–action space. We further observe positive transfer across algorithms and input modalities, including a transformer-based world-model. These results position flat-minima training as a simple, general mechanism for more robust MBRL without architectural changes.

ICRA Conference 2025 Conference Paper

Planning with Adaptive World Models for Autonomous Driving

• Arun Balajee Vasudevan
• Neehar Peri
• Jeff Schneider
• Deva Ramanan

Motion planning is crucial for safe navigation in complex urban environments. Historically, motion planners (MPs) have been evaluated with procedurally-generated simulators like CARLA. However, such synthetic benchmarks do not capture real-world multi-agent interactions. nuPlan, a recently released MP benchmark, addresses this limitation by augmenting real-world driving logs with closed-loop simulation logic, effectively turning the fixed dataset into a reactive simulator. We analyze the characteristics of nuPlan's recorded logs and find that each city has its own unique driving behaviors, suggesting that robust planners must adapt to different environments. We learn to model such unique behaviors with BehaviorNet, a graph convolutional neural network (GCNN) that predicts reactive agent behaviors using features derived from recently-observed agent histories; intuitively, some aggressive agents may tailgate lead vehicles, while others may not. To model such phenomena, BehaviorNet predicts the parameters of an agent's motion controller rather than directly predicting its spacetime trajectory (as most forecasters do). Finally, we present AdaptiveDriver, a model-predictive control (MPC) based planner that unrolls different world models conditioned on Behavior-Net's predictions. Our extensive experiments demonstrate that AdaptiveDriver achieves state-of-the-art results on the nuPlan closed-loop planning benchmark, improving over prior work by 2% on Test-14 Hard R-CLS, and generalizes even when evaluated on never-before-seen cities. project page

NeurIPS Conference 2025 Conference Paper

Retrospective In-Context Learning for Temporal Credit Assignment with Large Language Models

• Wen-Tse Chen
• Jiayu Chen
• Fahim Tajwar
• Hao Zhu
• Xintong Duan
• Ruslan Salakhutdinov
• Jeff Schneider

Learning from self-sampled data and sparse environmental feedback remains a fundamental challenge in training self-evolving agents. Temporal credit assignment mitigates this issue by transforming sparse feedback into dense supervision signals. However, previous approaches typically depend on domain-specific value functions for credit assignment, which suffer from poor sample efficiency and limited generalization. In this work, we propose to leverage pre-trained knowledge from large language models (LLMs) to transform sparse rewards into dense training signals (i. e. , the advantage function) through retrospective in-context learning (RICL). We further propose an online learning framework, RICOL, which iteratively refines the policy based on the credit assignment results from RICL. We empirically demonstrate that RICL can accurately estimate the advantage function with limited samples and effectively identify critical states for temporal credit assignment. Extended evaluation on the BabyAI benchmark shows that RICOL significantly improves sample efficiency compared to traditional online RL algorithms while achieving performance comparable to imitation learning from expert demonstartions. Our findings highlight the potential of leveraging LLMs for temporal credit assignment, paving the way for more sample-efficient and generalizable RL paradigms.

TMLR Journal 2025 Journal Article

State Combinatorial Generalization In Decision Making With Conditional Diffusion Models

• Xintong Duan
• Yutong He
• Fahim Tajwar
• Wentse Chen
• Ruslan Salakhutdinov
• Jeff Schneider

Many real-world decision-making problems are combinatorial in nature, where states (e.g., surrounding traffic of a self-driving car) can be seen as a combination of basic elements (e.g., pedestrians, trees, and other cars). Due to combinatorial complexity, observing all combinations of basic elements in the training set is infeasible, which leads to an essential yet understudied problem of zero-shot generalization to states that are unseen combinations of previously seen elements. In this work, we first formalize this problem and then demonstrate how existing value-based reinforcement learning (RL) algorithms struggle due to unreliable value predictions in unseen states. We argue that this problem cannot be addressed with exploration alone, but requires more expressive and generalizable models. We demonstrate that behavior cloning with a conditioned diffusion model trained on successful trajectory generalizes better to states formed by new combinations of seen elements than traditional RL methods. Through experiments in maze, driving, and multiagent environments, we show that conditioned diffusion models outperform traditional RL techniques and highlight the broad applicability of our problem formulation.

NeurIPS Conference 2025 Conference Paper

What is Your Data Worth to GPT? LLM-Scale Data Valuation with Influence Functions

• Sang Choe
• Hwijeen Ahn
• Juhan Bae
• Kewen Zhao
• Youngseog Chung
• Adithya Pratapa
• Willie Neiswanger
• Emma Strubell

Large language models (LLMs) are trained on a vast amount of human-written data, but data providers often remain uncredited. In response to this issue, data valuation (or data attribution), which quantifies the contribution or value of each data to the model output, has been discussed as a potential solution. Nevertheless, applying existing data valuation methods to recent LLMs and their vast training datasets has been largely limited by prohibitive compute and memory costs. In this work, we focus on influence functions, a popular gradient-based data valuation method, and significantly improve its scalability with an efficient gradient projection strategy called LoGra that leverages the gradient structure in backpropagation. We then provide a theoretical motivation of gradient projection approaches to influence functions to promote trust in the data valuation process. Lastly, we lower the barrier to implementing data valuation systems by introducing LogIX, a software package that can transform existing training code into data valuation code with minimal effort. In our data valuation experiments, LoGra achieves competitive accuracy against more expensive baselines while showing up to 6, 500x improvement in throughput and 5x reduction in GPU memory usage when applied to Llama3-8B-Instruct and the 1B-token dataset.

RLC Conference 2024 Conference Paper

Assigning Credit with Partial Reward Decoupling in Multi-Agent Proximal Policy Optimization

• Aditya Kapoor
• Benjamin Freed
• Jeff Schneider
• Howie Choset

Multi-agent proximal policy optimization (MAPPO) has recently demonstrated state-of-the-art performance on challenging multi-agent reinforcement learning tasks. However, MAPPO still struggles with the credit assignment problem, wherein the sheer difficulty in ascribing credit to individual agents' actions scales poorly with team size. In this paper, we propose a multi-agent reinforcement learning algorithm that adapts recent developments in credit assignment to improve upon MAPPO. Our approach leverages partial reward decoupling (PRD), which uses a learned attention mechanism to estimate which of a particular agent's teammates are relevant to its learning updates. We use this estimate to dynamically decompose large groups of agents into smaller, more manageable subgroups. We empirically demonstrate that our approach, PRD-MAPPO, decouples agents from teammates that do not influence their expected future reward, thereby streamlining credit assignment. We additionally show that PRD-MAPPO yields significantly higher data efficiency and asymptotic performance compared to both MAPPO and other state-of-the-art methods across several multi-agent tasks, including StarCraft II. Finally, we propose a version of PRD-MAPPO that is applicable to \textit{shared} reward settings, where PRD was previously not applicable, and empirically show that this also leads to performance improvements over MAPPO.

RLJ Journal 2024 Journal Article

Assigning Credit with Partial Reward Decoupling in Multi-Agent Proximal Policy Optimization

• Aditya Kapoor
• Benjamin Freed
• Jeff Schneider
• Howie Choset

Multi-agent proximal policy optimization (MAPPO) has recently demonstrated state-of-the-art performance on challenging multi-agent reinforcement learning tasks. However, MAPPO still struggles with the credit assignment problem, wherein the sheer difficulty in ascribing credit to individual agents' actions scales poorly with team size. In this paper, we propose a multi-agent reinforcement learning algorithm that adapts recent developments in credit assignment to improve upon MAPPO. Our approach leverages partial reward decoupling (PRD), which uses a learned attention mechanism to estimate which of a particular agent's teammates are relevant to its learning updates. We use this estimate to dynamically decompose large groups of agents into smaller, more manageable subgroups. We empirically demonstrate that our approach, PRD-MAPPO, decouples agents from teammates that do not influence their expected future reward, thereby streamlining credit assignment. We additionally show that PRD-MAPPO yields significantly higher data efficiency and asymptotic performance compared to both MAPPO and other state-of-the-art methods across several multi-agent tasks, including StarCraft II. Finally, we propose a version of PRD-MAPPO that is applicable to \textit{shared} reward settings, where PRD was previously not applicable, and empirically show that this also leads to performance improvements over MAPPO.

RLJ Journal 2024 Journal Article

Unifying Model-Based and Model-Free Reinforcement Learning with Equivalent Policy Sets

• Benjamin Freed
• Thomas Wei
• Roberto Calandra
• Jeff Schneider
• Howie Choset

Model-based and model-free reinforcement learning (RL) each possess relative strengths that prevent either algorithm from strictly outperforming the other. Model-based RL often offers greater data efficiency, as it can use models to evaluate many possible behaviors before choosing one to enact. However, because models cannot perfectly represent complex environments, agents that rely too heavily on models may suffer from poor asymptotic performance. Model-free RL, on the other hand, avoids this problem at the expense of data efficiency. In this work, we seek a unified approach to RL that combines the strengths of both approaches. To this end, we introduce the concept of _equivalent policy sets_ (EPS), which quantify the limitations of models for the purposes of decision-making, _i.e._, action selection. Based on this concept, we propose _Unified RL_, a novel RL algorithm that uses models to constrain model-free RL to the set of policies that are not provably suboptimal, according to model-based bounds on policy performance. We demonstrate across a range of benchmarks that Unified RL effectively combines the relative strengths of both model-based and model-free RL, in that it achieves comparable data efficiency to model-based RL, while achieving asymptotic performance similar or superior to that of model-free RL. Additionally, we show that Unified RL often outperforms a number of existing state-of-the-art model-based and model-free RL algorithms, and _can learn effective policies in situations where either model-based or model-free RL alone fail_.

RLC Conference 2024 Conference Paper

Unifying Model-Based and Model-Free Reinforcement Learning with Equivalent Policy Sets

• Benjamin Freed
• Thomas Wei
• Roberto Calandra
• Jeff Schneider
• Howie Choset

Model-based and model-free reinforcement learning (RL) each possess relative strengths that prevent either algorithm from strictly outperforming the other. Model-based RL often offers greater data efficiency, as it can use models to evaluate many possible behaviors before choosing one to enact. However, because models cannot perfectly represent complex environments, agents that rely too heavily on models may suffer from poor asymptotic performance. Model-free RL, on the other hand, avoids this problem at the expense of data efficiency. In this work, we seek a unified approach to RL that combines the strengths of both approaches. To this end, we introduce the concept of _equivalent policy sets_ (EPS), which quantify the limitations of models for the purposes of decision-making, _i. e. _, action selection. Based on this concept, we propose _Unified RL_, a novel RL algorithm that uses models to constrain model-free RL to the set of policies that are not provably suboptimal, according to model-based bounds on policy performance. We demonstrate across a range of benchmarks that Unified RL effectively combines the relative strengths of both model-based and model-free RL, in that it achieves comparable data efficiency to model-based RL, while achieving asymptotic performance similar or superior to that of model-free RL. Additionally, we show that Unified RL often outperforms a number of existing state-of-the-art model-based and model-free RL algorithms, and _can learn effective policies in situations where either model-based or model-free RL alone fail_.

NeurIPS Conference 2023 Conference Paper

PID-Inspired Inductive Biases for Deep Reinforcement Learning in Partially Observable Control Tasks

• Ian Char
• Jeff Schneider

Deep reinforcement learning (RL) has shown immense potential for learning to control systems through data alone. However, one challenge deep RL faces is that the full state of the system is often not observable. When this is the case, the policy needs to leverage the history of observations to infer the current state. At the same time, differences between the training and testing environments makes it critical for the policy not to overfit to the sequence of observations it sees at training time. As such, there is an important balancing act between having the history encoder be flexible enough to extract relevant information, yet be robust to changes in the environment. To strike this balance, we look to the PID controller for inspiration. We assert the PID controller's success shows that only summing and differencing are needed to accumulate information over time for many control tasks. Following this principle, we propose two architectures for encoding history: one that directly uses PID features and another that extends these core ideas and can be used in arbitrary control tasks. When compared with prior approaches, our encoders produce policies that are often more robust and achieve better performance on a variety of tracking tasks. Going beyond tracking tasks, our policies achieve 1. 7x better performance on average over previous state-of-the-art methods on a suite of locomotion control tasks.

NeurIPS Conference 2022 Conference Paper

Exploration via Planning for Information about the Optimal Trajectory

• Viraj Mehta
• Ian Char
• Joseph Abbate
• Rory Conlin
• Mark Boyer
• Stefano Ermon
• Jeff Schneider
• Willie Neiswanger

Many potential applications of reinforcement learning (RL) are stymied by the large numbers of samples required to learn an effective policy. This is especially true when applying RL to real-world control tasks, e. g. in the sciences or robotics, where executing a policy in the environment is costly. In popular RL algorithms, agents typically explore either by adding stochasticity to a reward-maximizing policy or by attempting to gather maximal information about environment dynamics without taking the given task into account. In this work, we develop a method that allows us to plan for exploration while taking both the task and the current knowledge about the dynamics into account. The key insight to our approach is to plan an action sequence that maximizes the expected information gain about the optimal trajectory for the task at hand. We demonstrate that our method learns strong policies with 2x fewer samples than strong exploration baselines and 200x fewer samples than model free methods on a diverse set of low-to-medium dimensional control tasks in both the open-loop and closed-loop control settings.

NeurIPS Conference 2021 Conference Paper

Beyond Pinball Loss: Quantile Methods for Calibrated Uncertainty Quantification

• Youngseog Chung
• Willie Neiswanger
• Ian Char
• Jeff Schneider

Among the many ways of quantifying uncertainty in a regression setting, specifying the full quantile function is attractive, as quantiles are amenable to interpretation and evaluation. A model that predicts the true conditional quantiles for each input, at all quantile levels, presents a correct and efficient representation of the underlying uncertainty. To achieve this, many current quantile-based methods focus on optimizing the pinball loss. However, this loss restricts the scope of applicable regression models, limits the ability to target many desirable properties (e. g. calibration, sharpness, centered intervals), and may produce poor conditional quantiles. In this work, we develop new quantile methods that address these shortcomings. In particular, we propose methods that can apply to any class of regression model, select an explicit balance between calibration and sharpness, optimize for calibration of centered intervals, and produce more accurate conditional quantiles. We provide a thorough experimental evaluation of our methods, which includes a high dimensional uncertainty quantification task in nuclear fusion.

ICRA Conference 2020 Conference Paper

Deep Kinematic Models for Kinematically Feasible Vehicle Trajectory Predictions

• Henggang Cui
• Thi Nguyen
• Fang-Chieh Chou
• Tsung-Han Lin
• Jeff Schneider
• David Bradley
• Nemanja Djuric

Self-driving vehicles (SDVs) hold great potential for improving traffic safety and are poised to positively affect the quality of life of millions of people. To unlock this potential one of the critical aspects of the autonomous technology is understanding and predicting future movement of vehicles surrounding the SDV. This work presents a deep-learning- based method for kinematically feasible motion prediction of such traffic actors. Previous work did not explicitly encode vehicle kinematics and instead relied on the models to learn the constraints directly from the data, potentially resulting in kinematically infeasible, suboptimal trajectory predictions. To address this issue we propose a method that seamlessly combines ideas from the AI with physically grounded vehicle motion models. In this way we employ best of the both worlds, coupling powerful learning models with strong feasibility guarantees for their outputs. The proposed approach is general, being applicable to any type of learning method. Extensive experiments using deep convnets on real-world data strongly indicate its benefits, outperforming the existing state-of-the-art.

JMLR Journal 2020 Journal Article

Tuning Hyperparameters without Grad Students: Scalable and Robust Bayesian Optimisation with Dragonfly

• Kirthevasan Kandasamy
• Karun Raju Vysyaraju
• Willie Neiswanger
• Biswajit Paria
• Christopher R. Collins
• Jeff Schneider
• Barnabas Poczos
• Eric P. Xing

Bayesian Optimisation (BO) refers to a suite of techniques for global optimisation of expensive black box functions, which use introspective Bayesian models of the function to efficiently search for the optimum. While BO has been applied successfully in many applications, modern optimisation tasks usher in new challenges where conventional methods fail spectacularly. In this work, we present Dragonfly, an open source Python library for scalable and robust BO. Dragonfly incorporates multiple recently developed methods that allow BO to be applied in challenging real world settings; these include better methods for handling higher dimensional domains, methods for handling multi-fidelity evaluations when cheap approximations of an expensive function are available, methods for optimising over structured combinatorial spaces, such as the space of neural network architectures, and methods for handling parallel evaluations. Additionally, we develop new methodological improvements in BO for selecting the Bayesian model, selecting the acquisition function, and optimising over complex domains with different variable types and additional constraints. We compare Dragonfly to a suite of other packages and algorithms for global optimisation and demonstrate that when the above methods are integrated, they enable significant improvements in the performance of BO. The Dragonfly library is available at dragonfly.github.io. [abs] [ pdf ][ bib ] [ code ] &copy JMLR 2020. ( edit, beta )

JAIR Journal 2019 Journal Article

Multi-fidelity Gaussian Process Bandit Optimisation

• Kirthevasan Kandasamy
• Gautam Dasarathy
• Junier Oliva
• Jeff Schneider
• Barnabás Póczos

In many scientific and engineering applications, we are tasked with the maximisation of an expensive to evaluate black box function f. Traditional settings for this problem assume just the availability of this single function. However, in many cases, cheap approximations to f may be obtainable. For example, the expensive real world behaviour of a robot can be approximated by a cheap computer simulation. We can use these approximations to eliminate low function value regions cheaply and use the expensive evaluations of f in a small but promising region and speedily identify the optimum. We formalise this task as a multi-fidelity bandit problem where the target function and its approximations are sampled from a Gaussian process. We develop MF-GP-UCB, a novel method based on upper confidence bound techniques. In our theoretical analysis we demonstrate that it exhibits precisely the above behaviour and achieves better bounds on the regret than strategies which ignore multi-fidelity information. Empirically, MF-GP-UCB outperforms such naive strategies and other multi-fidelity methods on several synthetic and real experiments.

ICRA Conference 2019 Conference Paper

Multimodal Trajectory Predictions for Autonomous Driving using Deep Convolutional Networks

• Henggang Cui
• Vladan Radosavljevic
• Fang-Chieh Chou
• Tsung-Han Lin
• Thi Nguyen
• Tzu-Kuo Huang
• Jeff Schneider
• Nemanja Djuric

Autonomous driving presents one of the largest problems that the robotics and artificial intelligence communities are facing at the moment, both in terms of difficulty and potential societal impact. Self-driving vehicles (SDVs) are expected to prevent road accidents and save millions of lives while improving the livelihood and life quality of many more. However, despite large interest and a number of industry players working in the autonomous domain, there still remains more to be done in order to develop a system capable of operating at a level comparable to best human drivers. One reason for this is high uncertainty of traffic behavior and large number of situations that an SDV may encounter on the roads, making it very difficult to create a fully generalizable system. To ensure safe and efficient operations, an autonomous vehicle is required to account for this uncertainty and to anticipate a multitude of possible behaviors of traffic actors in its surrounding. We address this critical problem and present a method to predict multiple possible trajectories of actors while also estimating their probabilities. The method encodes each actor's surrounding context into a raster image, used as input by deep convolutional networks to automatically derive relevant features for the task. Following extensive offline evaluation and comparison to state-of-the-art baselines, the method was successfully tested on SDVs in closed-course tests.

NeurIPS Conference 2019 Conference Paper

Offline Contextual Bayesian Optimization

• Ian Char
• Youngseog Chung
• Willie Neiswanger
• Kirthevasan Kandasamy
• Andrew Nelson
• Mark Boyer
• Egemen Kolemen
• Jeff Schneider

In black-box optimization, an agent repeatedly chooses a configuration to test, so as to find an optimal configuration. In many practical problems of interest, one would like to optimize several systems, or tasks'', simultaneously; however, in most of these scenarios the current task is determined by nature. In this work, we explore the offline'' case in which one is able to bypass nature and choose the next task to evaluate (e. g. via a simulator). Because some tasks may be easier to optimize and others may be more critical, it is crucial to leverage algorithms that not only consider which configurations to try next, but also which tasks to make evaluations for. In this work, we describe a theoretically grounded Bayesian optimization method to tackle this problem. We also demonstrate that if the model of the reward structure does a poor job of capturing variation in difficulty between tasks, then algorithms that actively pick tasks for evaluation may end up doing more harm than good. Following this, we show how our approach can be used for real world applications in science and engineering, including optimizing tokamak controls for nuclear fusion.

NeurIPS Conference 2018 Conference Paper

Neural Architecture Search with Bayesian Optimisation and Optimal Transport

• Kirthevasan Kandasamy
• Willie Neiswanger
• Jeff Schneider
• Barnabas Poczos
• Eric Xing

Bayesian Optimisation (BO) refers to a class of methods for global optimisation of a function f which is only accessible via point evaluations. It is typically used in settings where f is expensive to evaluate. A common use case for BO in machine learning is model selection, where it is not possible to analytically model the generalisation performance of a statistical model, and we resort to noisy and expensive training and validation procedures to choose the best model. Conventional BO methods have focused on Euclidean and categorical domains, which, in the context of model selection, only permits tuning scalar hyper-parameters of machine learning algorithms. However, with the surge of interest in deep learning, there is an increasing demand to tune neural network architectures. In this work, we develop NASBOT, a Gaussian process based BO framework for neural architecture search. To accomplish this, we develop a distance metric in the space of neural network architectures which can be computed efficiently via an optimal transport program. This distance might be of independent interest to the deep learning community as it may find applications outside of BO. We demonstrate that NASBOT outperforms other alternatives for architecture search in several cross validation based model selection tasks on multi-layer perceptrons and convolutional neural networks.

AAMAS Conference 2017 Conference Paper

Active Optimization and Self Driving Cars

• Jeff Schneider

AAAI Conference 2017 Conference Paper

Active Search for Sparse Signals with Region Sensing

• Yifei Ma
• Roman Garnett
• Jeff Schneider

Autonomous systems can be used to search for sparse signals in a large space; e. g. , aerial robots can be deployed to localize threats, detect gas leaks, or respond to distress calls. Intuitively, search algorithms may increase efficiency by collecting aggregate measurements summarizing large contiguous regions. However, most existing search methods either ignore the possibility of such region observations (e. g. , Bayesian optimization and multi-armed bandits) or make strong assumptions about the sensing mechanism that allow each measurement to arbitrarily encode all signals in the entire environment (e. g. , compressive sensing). We propose an algorithm that actively collects data to search for sparse signals using only noisy measurements of the average values on rectangular regions (including single points), based on the greedy maximization of information gain. We analyze our algorithm in 1d and show that it requires Õ(n/μ2 +k2 ) measurements to recover all of k signal locations with small Bayes error, where μ and n are the signal strength and the size of the search space, respectively. We also show that active designs can be fundamentally more efficient than passive designs with region sensing, contrasting with the results of Arias-Castro, Candes, and Davenport (2013). We demonstrate the empirical performance of our algorithm on a search problem using satellite image data and in high dimensions.

AAAI Conference 2017 Conference Paper

Enabling Dark Energy Science with Deep Generative Models of Galaxy Images

• Siamak Ravanbakhsh
• Francois Lanusse
• Rachel Mandelbaum
• Jeff Schneider
• Barnabas Poczos

Understanding the nature of dark energy, the mysterious force driving the accelerated expansion of the Universe, is a major challenge of modern cosmology. The next generation of cosmological surveys, specifically designed to address this issue, rely on accurate measurements of the apparent shapes of distant galaxies. However, shape measurement methods suffer from various unavoidable biases and therefore will rely on a precise calibration to meet the accuracy requirements of the science analysis. This calibration process remains an open challenge as it requires large sets of high quality galaxy images. To this end, we study the application of deep conditional generative models in generating realistic galaxy images. In particular we consider variations on conditional variational autoencoder and introduce a new adversarial objective for training of conditional generative networks. Our results suggest a reliable alternative to the acquisition of expensive high quality observations for generating the calibration data needed by the next generation of cosmological surveys.

AIJ Journal 2017 Journal Article

Query efficient posterior estimation in scientific experiments via Bayesian active learning

• Kirthevasan Kandasamy
• Jeff Schneider
• Barnabás Póczos

A common problem in disciplines of applied Statistics research such as Astrostatistics is of estimating the posterior distribution of relevant parameters. Typically, the likelihoods for such models are computed via expensive experiments such as cosmological simulations of the universe. An urgent challenge in these research domains is to develop methods that can estimate the posterior with few likelihood evaluations. In this paper, we study active posterior estimation in a Bayesian setting when the likelihood is expensive to evaluate. Existing techniques for posterior estimation are based on generating samples representative of the posterior. Such methods do not consider efficiency in terms of likelihood evaluations. In order to be query efficient we treat posterior estimation in an active regression framework. We propose two myopic query strategies to choose where to evaluate the likelihood and implement them using Gaussian processes. Via experiments on a series of synthetic and real examples we demonstrate that our approach is significantly more query efficient than existing techniques and other heuristics for posterior estimation.

IJCAI Conference 2017 Conference Paper

Scaling Active Search using Linear Similarity Functions

• Sibi Venkatesan
• James K. Miller
• Jeff Schneider
• Artur Dubrawski

Active Search has become an increasingly useful tool in information retrieval problems where the goal is to discover as many target elements as possible using only limited label queries. With the advent of big data, there is a growing emphasis on the scalability of such techniques to handle very large and very complex datasets. In this paper, we consider the problem of Active Search where we are given a similarity function between data points. We look at an algorithm introduced by Wang et al. [Wang et al. , 2013] known as Active Search on Graphs and propose crucial modifications which allow it to scale significantly. Their approach selects points by minimizing an energy function over the graph induced by the similarity function on the data. Our modifications require the similarity function to be a dot-product between feature vectors of data points, equivalent to having a linear kernel for the adjacency matrix. With this, we are able to scale tremendously: for n data points, the original algorithm runs in O(n^2) time per iteration while ours runs in only O(nr + r^2) given r-dimensional features. We also describe a simple alternate approach using a weighted-neighbor predictor which also scales well. In our experiments, we show that our method is competitive with existing semi-supervised approaches. We also briefly discuss conditions under which our algorithm performs well.

NeurIPS Conference 2016 Conference Paper

Gaussian Process Bandit Optimisation with Multi-fidelity Evaluations

• Kirthevasan Kandasamy
• Gautam Dasarathy
• Junier Oliva
• Jeff Schneider
• Barnabas Poczos

In many scientific and engineering applications, we are tasked with the optimisation of an expensive to evaluate black box function $\func$. Traditional methods for this problem assume just the availability of this single function. However, in many cases, cheap approximations to $\func$ may be obtainable. For example, the expensive real world behaviour of a robot can be approximated by a cheap computer simulation. We can use these approximations to eliminate low function value regions cheaply and use the expensive evaluations of $\func$ in a small but promising region and speedily identify the optimum. We formalise this task as a \emph{multi-fidelity} bandit problem where the target function and its approximations are sampled from a Gaussian process. We develop \mfgpucb, a novel method based on upper confidence bound techniques. In our theoretical analysis we demonstrate that it exhibits precisely the above behaviour, and achieves better regret than strategies which ignore multi-fidelity information. \mfgpucbs outperforms such naive strategies and other multi-fidelity methods on several synthetic and real experiments.

AAAI Conference 2016 Conference Paper

Linear-Time Learning on Distributions with Approximate Kernel Embeddings

• Danica Sutherland
• Junier Oliva
• Barnabás Póczos
• Jeff Schneider

Many interesting machine learning problems are best posed by considering instances that are distributions, or sample sets drawn from distributions. Most previous work devoted to machine learning tasks with distributional inputs has done so through pairwise kernel evaluations between pdfs (or sample sets). While such an approach is fine for smaller datasets, the computation of an N × N Gram matrix is prohibitive in large datasets. Recent scalable estimators that work over pdfs have done so only with kernels that use Euclidean metrics, like the L2 distance. However, there are a myriad of other useful metrics available, such as total variation, Hellinger distance, and the Jensen-Shannon divergence. This work develops the first random features for pdfs whose dot product approximates kernels using these non-Euclidean metrics. These random features allow estimators to scale to large datasets by working in a primal space, without computing large Gram matrices. We provide an analysis of the approximation error in using our proposed random features, and show empirically the quality of our approximation both in estimating a Gram matrix and in solving learning tasks in real-world and synthetic data.

IJCAI Conference 2016 Conference Paper

Nonparametric Risk and Stability Analysis for Multi-Task Learning Problems

• Xuezhi Wang
• Junier B. Oliva
• Jeff Schneider
• Barnab
• aacute; s P
• oacute; czos

Multi-task learning attempts to simultaneously leverage data from multiple domains in order to estimate related functions on each domain. For example, a special case of multi-task learning, transfer learning, is often employed when one has a good estimate of a function on a source domain, but is unable to estimate a related function well on a target domain using only target data. Multi-task/transfer learning problems are usually solved by imposing some kind of "smooth" relationship among/between tasks. In this paper, we study how different smoothness assumptions on task relations affect the upper bounds of algorithms proposed for these problems under different settings. For general multi-task learning, we study a family of algorithms which utilize a reweighting matrix on task weights to capture the smooth relationship among tasks, which has many instantiations in existing literature. Furthermore, for multi-task learning in a transfer learning framework, we study the recently proposed algorithms for the "model shift", where the conditional distribution $P(Y|X)$ is allowed to change across tasks but the change is assumed to be smooth. In addition, we illustrate our results with experiments on both simulated and real data.

NeurIPS Conference 2016 Conference Paper

The Multi-fidelity Multi-armed Bandit

• Kirthevasan Kandasamy
• Gautam Dasarathy
• Barnabas Poczos
• Jeff Schneider

We study a variant of the classical stochastic $K$-armed bandit where observing the outcome of each arm is expensive, but cheap approximations to this outcome are available. For example, in online advertising the performance of an ad can be approximated by displaying it for shorter time periods or to narrower audiences. We formalise this task as a \emph{multi-fidelity} bandit, where, at each time step, the forecaster may choose to play an arm at any one of $M$ fidelities. The highest fidelity (desired outcome) expends cost $\costM$. The $m$\ssth fidelity (an approximation) expends $\costm < \costM$ and returns a biased estimate of the highest fidelity. We develop \mfucb, a novel upper confidence bound procedure for this setting and prove that it naturally adapts to the sequence of available approximations and costs thus attaining better regret than naive strategies which ignore the approximations. For instance, in the above online advertising example, \mfucbs would use the lower fidelities to quickly eliminate suboptimal ads and reserve the larger expensive experiments on a small set of promising candidates. We complement this result with a lower bound and show that \mfucbs is nearly optimal under certain conditions.

IJCAI Conference 2015 Conference Paper

IJCAI-15 Distinguished Paper Bayesian Active Learning for Posterior Estimation

• Kirthevasan Kandasamy
• Jeff Schneider
• Barnabas Poczos

This paper studies active posterior estimation in a Bayesian setting when the likelihood is expensive to evaluate. Existing techniques for posterior estimation are based on generating samples representative of the posterior. Such methods do not consider efficiency in terms of likelihood evaluations. In order to be query efficient we treat posterior estimation in an active regression framework. We propose two myopic query strategies to choose where to evaluate the likelihood and implement them using Gaussian processes. Via experiments on a series of synthetic and real examples we demonstrate that our approach is significantly more query efficient than existing techniques and other heuristics for posterior estimation.

NeurIPS Conference 2014 Conference Paper

Flexible Transfer Learning under Support and Model Shift

• Xuezhi Wang
• Jeff Schneider

Transfer learning algorithms are used when one has sufficient training data for one supervised learning task (the source/training domain) but only very limited training data for a second task (the target/test domain) that is similar but not identical to the first. Previous work on transfer learning has focused on relatively restricted settings, where specific parts of the model are considered to be carried over between tasks. Recent work on covariate shift focuses on matching the marginal distributions on observations $X$ across domains. Similarly, work on target/conditional shift focuses on matching marginal distributions on labels $Y$ and adjusting conditional distributions $P(X|Y)$, such that $P(X)$ can be matched across domains. However, covariate shift assumes that the support of test $P(X)$ is contained in the support of training $P(X)$, i. e. , the training set is richer than the test set. Target/conditional shift makes a similar assumption for $P(Y)$. Moreover, not much work on transfer learning has considered the case when a few labels in the test domain are available. Also little work has been done when all marginal and conditional distributions are allowed to change while the changes are smooth. In this paper, we consider a general case where both the support and the model change across domains. We transform both $X$ and $Y$ by a location-scale shift to achieve transfer between tasks. Since we allow more flexible transformations, the proposed method yields better results on both synthetic data and real-world data.

NeurIPS Conference 2013 Conference Paper

Learning Hidden Markov Models from Non-sequence Data via Tensor Decomposition

• Tzu-Kuo Huang
• Jeff Schneider

Learning dynamic models from observed data has been a central issue in many scientific studies or engineering tasks. The usual setting is that data are collected sequentially from trajectories of some dynamical system operation. In quite a few modern scientific modeling tasks, however, it turns out that reliable sequential data are rather difficult to gather, whereas out-of-order snapshots are much easier to obtain. Examples include the modeling of galaxies, chronic diseases such Alzheimer's, or certain biological processes. Existing methods for learning dynamic model from non-sequence data are mostly based on Expectation-Maximization, which involves non-convex optimization and is thus hard to analyze. Inspired by recent advances in spectral learning methods, we propose to study this problem from a different perspective: moment matching and spectral decomposition. Under that framework, we identify reasonable assumptions on the generative process of non-sequence data, and propose learning algorithms based on the tensor decomposition method \cite{anandkumar2012tensor} to \textit{provably} recover first-order Markov models and hidden Markov models. To the best of our knowledge, this is the first formal guarantee on learning from non-sequence data. Preliminary simulation results confirm our theoretical findings.

NeurIPS Conference 2013 Conference Paper

Σ-Optimality for Active Learning on Gaussian Random Fields

• Yifei Ma
• Roman Garnett
• Jeff Schneider

A common classifier for unlabeled nodes on undirected graphs uses label propagation from the labeled nodes, equivalent to the harmonic predictor on Gaussian random fields (GRFs). For active learning on GRFs, the commonly used V-optimality criterion queries nodes that reduce the L2 (regression) loss. V-optimality satisfies a submodularity property showing that greedy reduction produces a (1 − 1/e) globally optimal solution. However, L2 loss may not characterise the true nature of 0/1 loss in classification problems and thus may not be the best choice for active learning. We consider a new criterion we call Σ-optimality, which queries the node that minimizes the sum of the elements in the predictive covariance. Σ-optimality directly optimizes the risk of the surveying problem, which is to determine the proportion of nodes belonging to one class. In this paper we extend submodularity guarantees from V-optimality to Σ-optimality using properties specific to GRFs. We further show that GRFs satisfy the suppressor-free condition in addition to the conditional independence inherited from Markov random fields. We test Σ-optimality on real-world graphs with both synthetic and real data and show that it outperforms V-optimality and other related methods on classification.

NeurIPS Conference 2011 Conference Paper

Group Anomaly Detection using Flexible Genre Models

• Liang Xiong
• Barnabás Póczos
• Jeff Schneider

An important task in exploring and analyzing real-world data sets is to detect unusual and interesting phenomena. In this paper, we study the group anomaly detection problem. Unlike traditional anomaly detection research that focuses on data points, our goal is to discover anomalous aggregated behaviors of groups of points. For this purpose, we propose the Flexible Genre Model (FGM). FGM is designed to characterize data groups at both the point level and the group level so as to detect various types of group anomalies. We evaluate the effectiveness of FGM on both synthetic and real data sets including images and turbulence data, and show that it is superior to existing approaches in detecting group anomalies.

NeurIPS Conference 2011 Conference Paper

Learning Auto-regressive Models from Sequence and Non-sequence Data

• Tzu-Kuo Huang
• Jeff Schneider

Vector Auto-regressive models (VAR) are useful tools for analyzing time series data. In quite a few modern time series modelling tasks, the collection of reliable time series turns out to be a major challenge, either due to the slow progression of the dynamic process of interest, or inaccessibility of repetitive measurements of the same dynamic process over time. In those situations, however, we observe that it is often easier to collect a large amount of non-sequence samples, or snapshots of the dynamic process of interest. In this work, we assume a small amount of time series data are available, and propose methods to incorporate non-sequence data into penalized least-square estimation of VAR models. We consider non-sequence data as samples drawn from the stationary distribution of the underlying VAR model, and devise a novel penalization scheme based on the discrete-time Lyapunov equation concerning the covariance of the stationary distribution. Experiments on synthetic and video data demonstrate the effectiveness of the proposed methods.

NeurIPS Conference 2010 Conference Paper

Learning Multiple Tasks with a Sparse Matrix-Normal Penalty

• Yi Zhang
• Jeff Schneider

In this paper, we propose a matrix-variate normal penalty with sparse inverse covariances to couple multiple tasks. Learning multiple (parametric) models can be viewed as estimating a matrix of parameters, where rows and columns of the matrix correspond to tasks and features, respectively. Following the matrix-variate normal density, we design a penalty that decomposes the full covariance of matrix elements into the Kronecker product of row covariance and column covariance, which characterizes both task relatedness and feature representation. Several recently proposed methods are variants of the special cases of this formulation. To address the overfitting issue and select meaningful task and feature structures, we include sparse covariance selection into our matrix-normal regularization via L-1 penalties on task and feature inverse covariances. We empirically study the proposed method and compare with related models in two real-world problems: detecting landmines in multiple fields and recognizing faces between different subjects. Experimental results show that the proposed framework provides an effective and flexible way to model various different structures of multiple tasks.

NeurIPS Conference 2008 Conference Paper

Learning the Semantic Correlation: An Alternative Way to Gain from Unlabeled Text

• Yi Zhang
• Artur Dubrawski
• Jeff Schneider

In this paper, we address the question of what kind of knowledge is generally transferable from unlabeled text. We suggest and analyze the semantic correlation of words as a generally transferable structure of the language and propose a new method to learn this structure using an appropriately chosen latent variable model. This semantic correlation contains structural information of the language space and can be used to control the joint shrinkage of model parameters for any specific task in the same space through regularization. In an empirical study, we construct 190 different text classification tasks from a real-world benchmark, and the unlabeled documents are a mixture from all these tasks. We test the ability of various algorithms to use the mixed unlabeled text to enhance all classification tasks. Empirical results show that the proposed approach is a reliable and scalable method for semi-supervised learning, regardless of the source of unlabeled data, the specific task to be enhanced, and the prediction model used.

NeurIPS Conference 2005 Conference Paper

Active Learning For Identifying Function Threshold Boundaries

• Brent Bryan
• Robert Nichol
• Christopher Genovese
• Jeff Schneider
• Christopher Miller
• Larry Wasserman

We present an efficient algorithm to actively select queries for learning the boundaries separating a function domain into regions where the func- tion is above and below a given threshold. We develop experiment selec- tion methods based on entropy, misclassification rate, variance, and their combinations, and show how they perform on a number of data sets. We then show how these algorithms are used to determine simultaneously valid 1 − α confidence intervals for seven cosmological parameters. Ex- perimentation shows that the algorithm reduces the computation neces- sary for the parameter estimation problem by an order of magnitude.

AAMAS Conference 2004 Conference Paper

Approximate Solutions for Partially Observable Stochastic Games with Common Payoffs

• Rosemary Emery-Montemerlo
• Geoff Gordon
• Jeff Schneider
• Sebastian Thrun

IJCAI Conference 2003 Conference Paper

Covariant Policy Search

• J. Andrew Bagnell
• Jeff Schneider

We investigate the problem of non-covariant behavior of policy gradient reinforcement learning algorithms. The policy gradient approach is amenable to analysis by information geometric methods. This leads us to propose a natural metric on controller parameterization that results from considering the manifold of probability distributions over paths induced by a stochastic controller. Investigation of this approach leads to a covariant gradient ascent rule. Interesting properties of this rule are discussed, including its relation with actor-critic style reinforcement learning algorithms. The algorithms discussed here are computationally quite efficient and on some interesting problems lead to dramatic performance improvement over noncovariant rules.

NeurIPS Conference 2003 Conference Paper

Policy Search by Dynamic Programming

• J. Bagnell
• Sham Kakade
• Jeff Schneider
• Andrew Ng

We consider the policy search approach to reinforcement learning. We show that if a “baseline distribution” is given (indicating roughly how often we expect a good policy to visit each state), then we can derive a policy search algorithm that terminates in a finite number of steps, and for which we can provide non-trivial performance guarantees. We also demonstrate this algorithm on several grid-world POMDPs, a planar biped walking robot, and a double-pole balancing problem.

AAAI Conference 2002 Conference Paper

Stochastic Link and Group Detection

• Jeremy Kubica
• Jeff Schneider

Link detection and analysis has long been important in the social sciences and in the government intelligence community. A significant effort is focused on the structural and functional analysis of “known” networks. Similarly, the detection of individual links is important but is usually done with techniques that result in “known” links. More recently the internet and other sources have led to a flood of circumstantial data that provide probabilistic evidence of links. Co-occurrence in news articles and simultaneous travel to the same location are two examples. We propose a probabilistic model of link generation based on membership in groups. The model considers both observed link evidence and demographic information about the entities. The parameters of the model are learned via a maximum likelihood search. In this paper we describe the model and then show several heuristics that make the search tractable. We test our model and optimization methods on synthetic data sets with a known ground truth and a database of news articles.

NeurIPS Conference 1996 Conference Paper

Exploiting Model Uncertainty Estimates for Safe Dynamic Control Learning

• Jeff Schneider

Model learning combined with dynamic programming has been shown to be effective for learning control of continuous state dynamic systems. The simplest method assumes the learned model is correct and applies dynamic programming to it, but many approximators provide uncertainty estimates on the fit. How can they be exploited? This paper addresses the case where the system must be prevented from having catastrophic failures dur(cid: 173) ing learning. We propose a new algorithm adapted from the dual control literature and use Bayesian locally weighted regression models with dy(cid: 173) namic programming. A common reinforcement learning assumption is that aggressive exploration should be encouraged. This paper addresses the con(cid: 173) verse case in which the system has to reign in exploration. The algorithm is illustrated on a 4 dimensional simulated control problem.

NeurIPS Conference 1995 Conference Paper

Memory-based Stochastic Optimization

• Andrew Moore
• Jeff Schneider

In this paper we introduce new algorithms for optimizing noisy plants in which each experiment is very expensive. The algorithms build a global non-linear model of the expected output at the same time as using Bayesian linear regression analysis of locally weighted polynomial models. The local model answers queries about confi(cid: 173) dence, noise, gradient and Hessians, and use them to make auto(cid: 173) mated decisions similar to those made by a practitioner of Response Surface Methodology. The global and local models are combined naturally as a locally weighted regression. We examine the ques(cid: 173) tion of whether the global model can really help optimization, and we extend it to the case of time-varying functions. We compare the new algorithms with a highly tuned higher-order stochastic op(cid: 173) timization algorithm on randomly-generated functions and a sim(cid: 173) ulated manufacturing task. We note significant improvements in total regret, time to converge, and final solution quality.