The Profiled Vehicle Routing Problem (PVRP) extends the classical VRP by incorporating vehicle–client-specific preferences and constraints, reflecting real‑world requirements such as zone restrictions and service‑level preferences. While recent reinforcement‑learning solvers have shown promising performance, they require retraining for each new profile distribution, suffer from poor representation ability, and struggle to generalize to out‑of‑distribution instances. In this paper, we address these limitations by introducing Unified Solver for Profiled Routing (USPR), a novel framework that natively handles arbitrary profile types. USPR introduces on three key innovations: (i) Profile Embeddings (PE) to encode any combination of profile types; (ii) Multi‑Head Profiled Attention (MHPA), an attention mechanism that models rich interactions between vehicles and clients; (iii) Profile‑aware Score Reshaping (PSR), which dynamically adjusts decoder logits using profile scores to improve generalization. Empirical results on diverse PVRP benchmarks demonstrate that USPR achieves state‑of‑the‑art results among learning‑based methods while offering significant gains in flexibility and computational efficiency. We make our source code publicly available to foster future research.

Trajectory prediction is a crucial task in modeling human behavior, especially in safety-critical fields such as social robotics and autonomous vehicle navigation. Traditional heuristics based on handcrafted rules often lack accuracy, while recently proposed deep learning approaches suffer from computational cost, slow inference speed, lack of explainability, and generalization issues that limit their practical adoption in such environments. In this paper, we introduce TrajEvo, a framework that leverages Large Language Models (LLMs) to automatically design trajectory prediction heuristics. TrajEvo employs an evolutionary algorithm to generate and refine prediction heuristics from past trajectory data. We introduce a Cross-Generation Elite Sampling to promote population diversity and a Statistics Feedback Loop allowing the LLM to analyze alternative predictions. Our evaluations show TrajEvo outperforms previous heuristic methods on various real-world datasets, and remarkably outperforms both heuristics and deep learning methods when generalizing to an unseen real-world dataset. TrajEvo represents a first step toward automated design of fast, explainable, and generalizable trajectory prediction heuristics. We make our source code publicly available to foster future research.

The offline datasets for imitation learning (IL) in multi-agent games typically contain player trajectories exhibiting diverse strategies, which necessitate measures to prevent learning algorithms from acquiring undesirable behaviors. Learning representations for these trajectories is an effective approach to depicting the strategies employed by each demonstrator. However, existing learning strategies often require player identification or rely on strong assumptions, which are not appropriate for multi-agent games. Therefore, in this paper, we introduce the Strategy Representation for Imitation Learning (STRIL) framework, which (1) effectively learns strategy representations in multi-agent games, (2) estimates proposed indicators based on these representations, and (3) filters out sub-optimal data using the indicators. STRIL is a plug-in method that can be integrated into existing IL algorithms. We demonstrate the effectiveness of STRIL across competitive multi-agent scenarios, including Two-player Pong, Limit Texas Hold'em, and Connect Four. Our approach successfully acquires strategy representations and indicators, thereby identifying dominant trajectories and significantly enhancing existing IL performance across these environments.

We present the Generative Flow Ant Colony Sampler (GFACS), a novel meta-heuristic method that hierarchically combines amortized inference and parallel stochastic search. Our method first leverages Generative Flow Networks (GFlowNets) to amortize a multi-modal prior distribution over combinatorial solution space that encompasses both high-reward and diversified solutions. This prior is iteratively updated via parallel stochastic search in the spirit of Ant Colony Optimization (ACO), leading to the posterior distribution that generates near-optimal solutions. Extensive experiments across seven combinatorial optimization problems demonstrate GFACS’s promising performances.

Recent advances in text-to-image diffusion models have achieved impressive image generation capabilities. However, it remains challenging to control the generation process with desired properties (e.g., aesthetic quality, user intention), which can be expressed as black-box reward functions. In this paper, we focus on prompt adaptation, which refines the original prompt into model-preferred prompts to generate desired images. While prior work uses reinforcement learning (RL) to optimize prompts, we observe that applying RL often results in generating similar postfixes and deterministic behaviors. To this end, we introduce Prompt Adaptation with GFlowNets (PAG), a novel approach that frames prompt adaptation as a probabilistic inference problem. Our key insight is that leveraging Generative Flow Networks (GFlowNets) allows us to shift from reward max-imization to sampling from an unnormalized density function, enabling both high-quality and diverse prompt generation. However, we identify that a naive application of GFlowNets suffers from mode collapse and uncovers a previously overlooked phenomenon: the progressive loss of neural plasticity in the model, which is compounded by inefficient credit assignment in sequential prompt generation. To address this critical challenge, we develop a systematic approach in PAG with flow reactivation, reward-prioritized sampling, and reward decomposition for prompt adaptation. Extensive experiments validate that PAG successfully learns to sample effective and diverse prompts for text-to-image generation. We also show that PAG exhibits strong robustness across various reward functions and transferability to different text-to-image models.

Amortized inference is the task of training a parametric model, such as a neural network, to approximate a distribution with a given unnormalized density where exact sampling is intractable. When sampling is modeled as a sequential decision-making process, reinforcement learning (RL) methods, such as generative flow networks, can be used to train the sampling policy. Off-policy RL training facilitates the discovery of diverse, high-reward candidates, but existing methods still face challenges in efficient exploration. We propose to use an adaptive training distribution (the Teacher) to guide the training of the primary amortized sampler (the Student). The Teacher, an auxiliary behavior model, is trained to sample high-loss regions of the Student and can generalize across unexplored modes, thereby enhancing mode coverage by providing an efficient training curriculum. We validate the effectiveness of this approach in a synthetic environment designed to present an exploration challenge, two diffusion-based sampling tasks, and four biochemical discovery tasks demonstrating its ability to improve sample efficiency and mode coverage. Source code is available at https://github.com/alstn12088/adaptive-teacher.

Metal-organic frameworks (MOFs) are a class of crystalline materials with promising applications in many areas such as carbon capture and drug delivery. In this work, we introduce MOFFlow, the first deep generative model tailored for MOF structure prediction. Existing approaches, including ab initio calculations and even deep generative models, struggle with the complexity of MOF structures due to the large number of atoms in the unit cells. To address this limitation, we propose a novel Riemannian flow matching framework that reduces the dimensionality of the problem by treating the metal nodes and organic linkers as rigid bodies, capitalizing on the inherent modularity of MOFs. By operating in the $SE(3)$ space, MOFFlow effectively captures the roto-translational dynamics of these rigid components in a scalable way. Our experiment demonstrates that MOFFlow accurately predicts MOF structures containing several hundred atoms, significantly outperforming conventional methods and state-of-the-art machine learning baselines while being much faster. Code available at https://github.com/nayoung10/MOFFlow.

Generative models in drug discovery have recently gained attention as efficient alternatives to brute-force virtual screening. However, most existing models do not account for synthesizability, limiting their practical use in real-world scenarios. In this paper, we propose RxnFlow, which sequentially assembles molecules using predefined molecular building blocks and chemical reaction templates to constrain the synthetic chemical pathway. We then train on this sequential generating process with the objective of generative flow networks (GFlowNets) to generate both highly rewarded and diverse molecules. To mitigate the large action space of synthetic pathways in GFlowNets, we implement a novel action space subsampling method. This enables RxnFlow to learn generative flows over extensive action spaces comprising combinations of 1.2 million building blocks and 71 reaction templates without significant computational overhead. Additionally, RxnFlow can employ modified or expanded action spaces for generation without retraining, allowing for the introduction of additional objectives or the incorporation of newly discovered building blocks. We experimentally demonstrate that RxnFlow outperforms existing reaction-based and fragment-based models in pocket-specific optimization across various target pockets. Furthermore, RxnFlow achieves state-of-the-art performance on CrossDocked2020 for pocket-conditional generation, with an average Vina score of –8.85 kcal/mol and 34.8% synthesizability. Code is available at https://github.com/SeonghwanSeo/RxnFlow.

Designing biological sequences with desired properties is challenging due to vast search spaces and limited evaluation budgets. Although reinforcement learning methods use proxy models for rapid reward evaluation, insufficient training data can cause proxy misspecification on out-of-distribution inputs. To address this, we propose a novel off-policy search, $\delta$-Conservative Search, that enhances robustness by restricting policy exploration to reliable regions. Starting from high-score offline sequences, we inject noise by randomly masking tokens with probability $\delta$, then denoise them using our policy. We further adapt $\delta$ based on proxy uncertainty on each data point, aligning the level of conservativeness with model confidence. Experimental results show that our conservative search consistently enhances the off-policy training, outperforming existing machine learning methods in discovering high-score sequences across diverse tasks, including DNA, RNA, protein, and peptide design.

Effective search methods are crucial for improving the performance of deep generative models at test time. In this paper, we introduce a novel test-time search method, Neural Genetic Search (NGS), which incorporates the evolutionary mechanism of genetic algorithms into the generation procedure of deep models. The core idea behind NGS is its crossover, which is defined as parent-conditioned generation using trained generative models. This approach offers a versatile and easy-to-implement search algorithm for deep generative models. We demonstrate the effectiveness and flexibility of NGS through experiments across three distinct domains: routing problems, adversarial prompt generation for language models, and molecular design.

Optimizing high-dimensional and complex black-box functions is crucial in numerous scientific applications. While Bayesian optimization (BO) is a powerful method for sample-efficient optimization, it struggles with the curse of dimensionality and scaling to thousands of evaluations. Recently, leveraging generative models to solve black-box optimization problems has emerged as a promising framework. However, those methods often underperform compared to BO methods due to limited expressivity and difficulty of uncertainty estimation in high-dimensional spaces. To overcome these issues, we introduce DiBO, a novel framework for solving high-dimensional black-box optimization problems. Our method iterates two stages. First, we train a diffusion model to capture the data distribution and deep ensembles to predict function values with uncertainty quantification. Second, we cast the candidate selection as a posterior inference problem to balance exploration and exploitation in high-dimensional spaces. Concretely, we fine-tune diffusion models to amortize posterior inference. Extensive experiments demonstrate that our method outperforms state-of-the-art baselines across synthetic and real-world tasks. Our code is publicly available https://github.com/umkiyoung/DiBO.

Multi-Agent Reinforcement Learning (MARL) has shown promise in solving complex problems involving cooperation and competition among agents, such as an Unmanned Surface Vehicle (USV) swarm used in search and rescue, surveillance, and vessel protection. However, aligning system behavior with user preferences is challenging due to the difficulty of encoding expert intuition into reward functions. To address the issue, we propose a Reinforcement Learning with Human Feedback (RLHF) approach for MARL that resolves credit-assignment challenges through an Agent-Level Feedback system categorizing feedback into intra-agent, inter-agent, and intra-team types. To overcome the challenges of direct human feedback, we employ a Large Language Model (LLM) evaluator to validate our approach using feedback scenarios such as region constraints, collision avoidance, and task allocation. Our method effectively refines USV swarm policies, addressing key challenges in multi-agent systems while maintaining fairness and performance consistency.

Combinatorial optimization (CO) is fundamental to several realworld applications, from logistics and scheduling to hardware design and resource allocation.Deep reinforcement learning (RL) has recently shown significant benefits in solving CO problems, reducing reliance on domain expertise and improving computational efficiency.However, the absence of a unified benchmarking framework leads to inconsistent evaluations, limits reproducibility, and increases engineering overhead, raising barriers to adoption for new researchers.To address these challenges, we introduce RL4CO, a unified and extensive benchmark with in-depth library coverage of 27 CO problem environments and 23 state-of-the-art baselines.Built on efficient software libraries and best practices in implementation, RL4CO features modularized implementation and flexible configurations of diverse environments, policy architectures, RL algorithms, and utilities with extensive documentation.RL4CO helps researchers build on existing successes while exploring and developing their own designs, facilitating the entire research process by decoupling science from heavy engineering.We finally provide extensive benchmark studies to inspire new insights and future work.RL4CO has already attracted numerous researchers in the community and is open-sourced at https://github.com/ai4co/rl4co 1 .

Wind farms generate electricity from wind, offering a sustainable and eco-friendly power source.In wind farms, determining the positions of wind turbines is crucial for high energy production due to complex wake interactions.However, optimizing wind farm layouts for given wind conditions remains challenging.While existing methods try to solve the problem by training a surrogate model and employing optimization algorithms based on the model, these methods require a large number of simulations to obtain highly productive layouts, which is time-consuming in large-scale tasks.Furthermore, they mostly yield less diverse layouts, making the deployment of such layouts into real-world scenarios difficult.To address these challenges, we introduce a novel conditional generative modeling approach to find wind farm layouts that maximize energy production.Our method consists of four stages.First, we collect datasets that consist of layouts and their corresponding annual energy production (AEP).Then, we train a diffusion model conditioned on AEP and wind scenarios using the collected dataset.Specifically, we use a Graph Neural Network as the backbone for the denoising network to ensure permutation invariance.Next, we sample promising layouts from the trained diffusion model by conditioning with a high AEP and given wind scenario.We also introduce a local search algorithm, which enables us to adjust layouts that violate design constraints.Finally, we evaluate the generated layouts and augment the dataset.We repeat these processes iteratively to further optimize the layouts.Throughout these processes, we effectively generate highly productive and diverse wind farm layouts in a sample-efficient manner.Experimental results demonstrate that our approach outperforms state-of-the-art methods in maximizing energy production across various scenarios.

Combinatorial optimization problems involving multiple agents are notoriously challenging due to their NP-hard nature and the necessity for effective agent coordination. Despite advancements in learning-based methods, existing approaches often face critical limitations, including suboptimal agent coordination, poor generalization, and high computational latency. To address these issues, we propose PARCO (Parallel AutoRegressive Combinatorial Optimization), a general reinforcement learning framework designed to construct high-quality solutions for multi-agent combinatorial tasks efficiently. To this end, PARCO integrates three key novel components: (1) transformer-based communication layers to enable effective agent collaboration during parallel solution construction, (2) a multiple pointer mechanism for low-latency, parallel agent decision-making, and (3) priority-based conflict handlers to resolve decision conflicts via learned priorities. We evaluate PARCO in multi-agent vehicle routing and scheduling problems, where our approach outperforms state-of-the-art learning methods, demonstrating strong generalization ability and remarkable computational efficiency. We make our source code publicly available to foster future research: https://github.com/ai4co/parco.

Reinforcement learning (RL) has emerged as an effective approach for signal control, and several studies attempt to apply RL to network-wide signal control problems. Most existing methods for network-wide signal controls simply apply the pretrained model independently to the network or simply expand the observation range for the neighbors. While several recent studies have proposed cooperative signal control models using multi-agent reinforcement learning (MARL) approaches, they have not considered the transferability of the policy. This issue becomes more critical in treating a multi-agent signal control problem, since there exist innumerable possibilities of data distributions for the demand patterns. Hence, this research aims to develop a transferable and cooperative multi-agent traffic signal control model. As the key idea, we propose a decentralized and communication-based MARL approach that considers the spatial correlation of traffic dynamics in the model design and employs a graph representation for the state. Three experiments in simulated environments of the Gangnam district in Seoul, South Korea are conducted to 1) validate the model, 2) assess the policy’s transferability, and 3) evaluate the efficiency of the multi-agent cooperation. The experimental results show that the proposed model obtains a transferable policy so that it adapts to the unexperienced demand scenario. In addition, it is concluded that the joint action of the proposed RL model cooperatively rebalances the traffic demands so that it improves the efficiency of the network-wide signal controls.

Min-max routing problems aim to minimize the maximum tour length among multiple agents as they collaboratively visit all cities, i.e., the completion time. These problems include impactful real-world applications but are known as NP-hard. Existing methods are facing challenges, particularly in large-scale problems that require the coordination of numerous agents to cover thousands of cities. This paper proposes Equity-Transformer to solve large-scale min-max routing problems. First, we model min-max routing problems into sequential planning, reducing the complexity and enabling the use of a powerful Transformer architecture. Second, we propose key inductive biases that ensure equitable workload distribution among agents. The effectiveness of Equity-Transformer is demonstrated through its superior performance in two representative min-max routing tasks: the min-max multi-agent traveling salesman problem (min-max mTSP) and the min-max multi-agent pick-up and delivery problem (min-max mPDP). Notably, our method achieves significant reductions of runtime, approximately 335 times, and cost values of about 53% compared to a competitive heuristic (LKH3) in the case of 100 vehicles with 1,000 cities of mTSP. We provide reproducible source code: https://github.com/kaist-silab/equity-transformer.

Generative Flow Networks (GFlowNets) are amortized sampling methods that learn a distribution over discrete objects proportional to their rewards. GFlowNets exhibit a remarkable ability to generate diverse samples, yet occasionally struggle to consistently produce samples with high rewards due to over-exploration on wide sample space. This paper proposes to train GFlowNets with local search, which focuses on exploiting high-rewarded sample space to resolve this issue. Our main idea is to explore the local neighborhood via backtracking and reconstruction guided by backward and forward policies, respectively. This allows biasing the samples toward high-reward solutions, which is not possible for a typical GFlowNet solution generation scheme, which uses the forward policy to generate the solution from scratch. Extensive experiments demonstrate a remarkable performance improvement in several biochemical tasks. Source code is available: \url{https://github.com/dbsxodud-11/ls_gfn}.

Deep reinforcement learning (DRL) has significantly advanced the field of combinatorial optimization (CO). However, its practicality is hindered by the necessity for a large number of reward evaluations, especially in scenarios involving computationally intensive function assessments. To enhance the sample efficiency, we propose a simple but effective method, called symmetric replay training (SRT), which can be easily integrated into various DRL methods. Our method leverages high-reward samples to encourage exploration of the under-explored symmetric regions without additional online interactions - free. Through replay training, the policy is trained to maximize the likelihood of the symmetric trajectories of discovered high-rewarded samples. Experimental results demonstrate the consistent improvement of our method in sample efficiency across diverse DRL methods applied to real-world tasks, such as molecular optimization and hardware design.

GFlowNets are probabilistic models that sequentially generate compositional structures through a stochastic policy. Among GFlowNets, temperature-conditional GFlowNets can introduce temperature-based controllability for exploration and exploitation. We propose Logit-scaling GFlowNets (Logit-GFN), a novel architectural design that greatly accelerates the training of temperature-conditional GFlowNets. It is based on the idea that previously proposed approaches introduced numerical challenges in the deep network training, since different temperatures may give rise to very different gradient profiles as well as magnitudes of the policy’s logits. We find that the challenge is greatly reduced if a learned function of the temperature is used to scale the policy’s logits directly. Also, using Logit-GFN, GFlowNets can be improved by having better generalization capabilities in offline learning and mode discovery capabilities in online learning, which is empirically verified in various biological and chemical tasks. Our code is available at https://github.com/dbsxodud-11/logit-gfn

In this work, we propose a novel method, NLNS-MASPF, to solve the Multi-Agent Scheduling and Pathfinding (MASPF) problem. The problem exhibits a bi-level structure, consisting of High-level Scheduling and Low-level Pathfinding. Our method applies a graph neural network in the high-level scheduling process and utilizes a MAPF solver with a schedule segmenting technique in the low-level pathfinding process. Through these approaches, NLNS-MASPF has experimentally demonstrated superior performance compared to the previous state-of-the-art MASPF algorithm, LNS-PBS, in solving the MASPF problem.

Multi-Agent Reinforcement Learning (MARL) based Multi-Agent Path Finding (MAPF) has recently gained attention due to its efficiency and scalability. Several MARL-MAPF methods choose to use communication to enrich the information one agent can perceive. However, existing works still struggle in structured environments with high obstacle density and a high number of agents. To further improve the performance of the communication-based MARL-MAPF solvers, we propose a new method, Ensembling Prioritized Hybrid Policies (EPH). We first propose a selective communication block to gather richer information for better agent coordination within multi-agent environments and train the model with a Q-learning-based algorithm. We further introduce three advanced inference strategies aimed at bolstering performance during the execution phase. First, we hybridize the neural policy with single-agent expert guidance for navigating conflict-free zones. Secondly, we propose Q value-based methods for prioritized resolution of conflicts as well as deadlock situations. Finally, we introduce a robust ensemble method that can efficiently collect the best out of multiple possible solutions. We empirically evaluate EPH in complex multi-agent environments and demonstrate competitive performance against state-of-the-art neural methods for MAPF. We open-source our code at https://github.com/ai4co/eph-mapf.

Complex urban road networks with high vehicle occupancy frequently face severe traffic congestion. Designing an effective strategy for managing multiple traffic lights plays a crucial role in managing congestion. However, most current traffic light management systems rely on human-crafted decisions, which may not adapt well to diverse traffic patterns. In this paper, we delve into two pivotal design components of the traffic light management system that can be dynamically adjusted to various traffic conditions: phase combination and phase time allocation. While numerous studies have sought an efficient strategy for managing traffic lights, most of these approaches consider a fixed traffic pattern and are limited to relatively small road networks. To overcome these limitations, we introduce a novel and practical framework to formulate the optimization of such design components using an offline meta black-box optimization. We then present a simple yet effective method to efficiently find a solution for the aforementioned problem. In our framework, we first collect an offline meta dataset consisting of pairs of design choices and corresponding congestion measures from various traffic patterns. After collecting the dataset, we employ the Attentive Neural Process (ANP) to predict the impact of the proposed design on congestion across various traffic patterns with well-calibrated uncertainty. Finally, Bayesian optimization, with ANP as a surrogate model, is utilized to find an optimal design for unseen traffic patterns through limited online simulations. Our experiment results show that our method outperforms state-of-the-art baselines on complex road networks in terms of the number of waiting vehicles. Surprisingly, the deployment of our method into a real-world traffic system was able to improve traffic throughput by 4.80% compared to the original strategy.

This paper studies Generative Flow Networks (GFlowNets), which learn to sample objects proportionally to a given reward function through the trajectory of state transitions. In this work, we observe that GFlowNets tend to under-exploit the high-reward objects due to training on insufficient number of trajectories, which may lead to a large gap between the estimated flow and the (known) reward value. In response to this challenge, we propose a pessimistic backward policy for GFlowNets (PBP-GFN), which maximizes the observed flow to align closely with the true reward for the object. We extensively evaluate PBP-GFN across eight benchmarks, including hyper-grid environment, bag generation, structured set generation, molecular generation, and four RNA sequence generation tasks. In particular, PBP-GFN enhances the discovery of high-reward objects, maintains the diversity of the objects, and consistently outperforms existing methods.

The challenge of discovering new molecules with desired properties is crucial in domains like drug discovery and material design. Recent advances in deep learning-based generative methods have shown promise but face the issue of sample efficiency due to the computational expense of evaluating the reward function. This paper proposes a novel algorithm for sample-efficient molecular optimization by distilling a powerful genetic algorithm into deep generative policy using GFlowNets training, the off-policy method for amortized inference. This approach enables the deep generative policy to learn from domain knowledge, which has been explicitly integrated into the genetic algorithm. Our method achieves state-of-the-art performance in the official molecular optimization benchmark, significantly outperforming previous methods. It also demonstrates effectiveness in designing inhibitors against SARS-CoV-2 with substantially fewer reward calls.

Offline Reinforcement Learning (Offline RL) presents challenges of learning effective decision-making policies from static datasets without any online interactions. Data augmentation techniques, such as noise injection and data synthesizing, aim to improve Q-function approximation by smoothing the learned state-action region. However, these methods often fall short of directly improving the quality of offline datasets, leading to suboptimal results. In response, we introduce GTA, Generative Trajectory Augmentation, a novel generative data augmentation approach designed to enrich offline data by augmenting trajectories to be both high-rewarding and dynamically plausible. GTA applies a diffusion model within the data augmentation framework. GTA partially noises original trajectories and then denoises them with classifier-free guidance via conditioning on amplified return value. Our results show that GTA, as a general data augmentation strategy, enhances the performance of widely used offline RL algorithms across various tasks with unique challenges. Furthermore, we conduct a quality analysis of data augmented by GTA and demonstrate that GTA improves the quality of the data. Our code is available at https://github.com/Jaewoopudding/GTA

The omnipresence of NP-hard combinatorial optimization problems (COPs) compels domain experts to engage in trial-and-error heuristic design. The long-standing endeavor of design automation has gained new momentum with the rise of large language models (LLMs). This paper introduces Language Hyper-Heuristics (LHHs), an emerging variant of Hyper-Heuristics that leverages LLMs for heuristic generation, featuring minimal manual intervention and open-ended heuristic spaces. To empower LHHs, we present Reflective Evolution (ReEvo), a novel integration of evolutionary search for efficiently exploring the heuristic space, and LLM reflections to provide verbal gradients within the space. Across five heterogeneous algorithmic types, six different COPs, and both white-box and black-box views of COPs, ReEvo yields state-of-the-art and competitive meta-heuristics, evolutionary algorithms, heuristics, and neural solvers, while being more sample-efficient than prior LHHs.

Optimizing complex and high-dimensional black-box functions is ubiquitous in science and engineering fields. Unfortunately, the online evaluation of these functions is restricted due to time and safety constraints in most cases. In offline model-based optimization (MBO), we aim to find a design that maximizes the target function using only a pre-existing offline dataset. While prior methods consider forward or inverse approaches to address the problem, these approaches are limited by conservatism and the difficulty of learning highly multi-modal mappings. Recently, there has been an emerging paradigm of learning to improve solutions with synthetic trajectories constructed from the offline dataset. In this paper, we introduce a novel conditional generative modeling approach to produce trajectories toward high-scoring regions. First, we construct synthetic trajectories toward high-scoring regions using the dataset while injecting locality bias for consistent improvement directions. Then, we train a conditional diffusion model to generate trajectories conditioned on their scores. Lastly, we sample multiple trajectories from the trained model with guidance to explore high-scoring regions beyond the dataset and select high-fidelity designs among generated trajectories with the proxy function. Extensive experiment results demonstrate that our method outperforms competitive baselines on Design-Bench and its practical variants. The code is publicly available in \url{https://github.com/dbsxodud-11/GTG}.

The cutting plane method is a key technique for successful branch-and-cut and branch-price-and-cut algorithms that find the exact optimal solutions for various vehicle routing problems (VRPs). Among various cuts, the rounded capacity inequalities (RCIs) are the most fundamental. To generate RCIs, we need to solve the separation problem, whose exact solution takes a long time to obtain; therefore, heuristic methods are widely used. We design a learning-based separation heuristic algorithm with graph coarsening that learns the solutions of the exact separation problem with a graph neural network (GNN), which is trained with small instances of 50 to 100 customers. We embed our separation algorithm within the cutting plane method to find a lower bound for the capacitated VRP (CVRP) with up to 1,000 customers. We compare the performance of our approach with CVRPSEP, a popular separation software package for various cuts used in solving VRPs. Our computational results show that our approach finds better lower bounds than CVRPSEP for large-scale problems with 400 or more customers, whereas CVRPSEP shows strong competency for problems with less than 400 customers. History: Accepted by Andrea Lodi, Area Editor for Design & Analysis of Algorithms–Discrete. Funding: This work was supported by the National Research Foundation of Korea (NRF) grant funded by the Korean government (MSIT) [RS-2023-00259550] and the Institute of Information & Communications Technology Planning & Evaluation (IITP) grant funded by the Korean government (MSIT) [2022-0-01032, Development of Collective Collaboration Intelligence Framework for Internet of Autonomous Things]. Supplemental Material: The online appendix is available at https://doi.org/10.1287/ijoc.2022.0310 .

Deep reinforcement learning (DRL) provides a promising way for intelligent agents (e.g., autonomous vehicles) to learn to navigate complex scenarios. However, DRL with neural networks as function approximators is typically considered a black box with little explainability and often suffers from suboptimal performance, especially for autonomous navigation in highly interactive multi-agent environments. To address these issues, we propose three auxiliary tasks with spatio-temporal relational reasoning and integrate them into the standard DRL framework, which improves the decision making performance and provides explainable intermediate indicators. We propose to explicitly infer the internal states (i.e., traits and intentions) of surrounding agents (e.g., human drivers) as well as to predict their future trajectories in the situations with and without the ego agent through counterfactual reasoning. These auxiliary tasks provide additional supervision signals to infer the behavior patterns of other interactive agents. Multiple variants of framework integration strategies are compared. We also employ a spatio-temporal graph neural network to encode relations between dynamic entities, which enhances both internal state inference and decision making of the ego agent. Moreover, we propose an interactivity estimation mechanism based on the difference between predicted trajectories in these two situations, which indicates the degree of influence of the ego agent on other agents. To validate the proposed method, we design an intersection driving simulator based on the Intelligent Intersection Driver Model (IIDM) that simulates vehicles and pedestrians. Our approach achieves robust and state-of-the-art performance in terms of standard evaluation metrics and provides explainable intermediate indicators (i.e., internal states, and interactivity scores) for decision making.

Convolutional deep sets is a neural network architecture that can model stationary stochastic processes. This architecture uses the kernel smoother and deep convolutional neural network to construct translation equivariant functional representations. However, the non-parametric nature of the kernel smoother can produce ambiguous representations when the number of data points is not given sufficiently. To address this issue, we introduce bayesian convolutional deep sets, which constructs random translation equivariant functional representations with a stationary prior. Furthermore, we present how to impose the task-dependent prior for each dataset because a wrongly imposed prior can result in an even worse representation than that of the kernel smoother. Empirically, we demonstrate that the proposed architecture alleviates the targeted issue in various experiments with time-series and image datasets.

In this paper, we present DevFormer, a novel transformer-based architecture for addressing the complex and computationally demanding problem of hardware design optimization. Despite the demonstrated efficacy of transformers in domains including natural language processing and computer vision, their use in hardware design has been limited by the scarcity of offline data. Our approach addresses this limitation by introducing strong inductive biases such as relative positional embeddings and action-permutation symmetricity that effectively capture the hardware context and enable efficient design optimization with limited offline data. We apply DevFormer to the problem of decoupling capacitor placement and show that it outperforms state-of-the-art methods in both simulated and real hardware, leading to improved performances while reducing the number of components by more than 30%. Finally, we show that our approach achieves promising results in other offline contextual learning-based combinatorial optimization tasks.

This paper proposes Meta-SAGE, a novel approach for improving the scalability of deep reinforcement learning models for combinatorial optimization (CO) tasks. Our method adapts pre-trained models to larger-scale problems in test time by suggesting two components: a scale meta-learner (SML) and scheduled adaptation with guided exploration (SAGE). First, SML transforms the context embedding for subsequent adaptation of SAGE based on scale information. Then, SAGE adjusts the model parameters dedicated to the context embedding for a specific instance. SAGE introduces locality bias, which encourages selecting nearby locations to determine the next location. The locality bias gradually decays as the model is adapted to the target instance. Results show that Meta-SAGE outperforms previous adaptation methods and significantly improves scalability in representative CO tasks. Our source code is available at https://github.com/kaist-silab/meta-sage.

In this work, we consider the problem of allocating tasks and planning paths for multiple robots to operate cooper-atively. We formulate the problem as a bi-level optimization that involves optimizing the scheduling of robots and the collision-free path for each robot. To address the complexity of the environment with obstacles, we introduce a congestion-aware state representation technique with the aid of graph neural networks. We also derive a joint scheduling policy using multi-agent reinforcement learning, and we propose an additional auxiliary loss that promotes coordination among the robots. Through empirical evaluation, we demonstrate the effectiveness of our approach in solving the combined task allocation and path planning problem in a cooperative multi-robot system.

This paper presents a novel approach for modeling and predicting patterns of events in time-series learning, named graph recurrent temporal point process (GRTPP). Prior research has focused on using deep learning techniques, such as recurrent neural networks (RNNs) or attention-based sequential data embedding, on modeling the time-varying intensity of events. However, these models were typically limited to modeling a single intensity function capturing the event occurrence of all event types simultaneously. GRTPP addresses this issue by encoding multivariate event sequences into a sequence of graphs, where each node contains information about the event occurrence and time. The sequence of graphs is then embedded into node embeddings for each event type, taking into account the relationships between the event types. By integrating the estimated intensity functions, GRTPP predicts the event type and the timing of the next event. The proposed GRTPP model offers improved effectiveness and explainability compared to previous models, as demonstrated through empirical evaluations on five real-world datasets and the actual credit card transaction dataset. The code is available at https://github.com/im0j/GRTPP https://github.com/im0j/GRTPP.

While complex simulations of physical systems have been widely used in engineering and scientific computing, lowering their often prohibitive computational requirements has only recently been tackled by deep learning approaches. In this paper, we present GraphSplineNets, a novel deep-learning method to speed up the forecasting of physical systems by reducing the grid size and number of iteration steps of deep surrogate models. Our method uses two differentiable orthogonal spline collocation methods to efficiently predict response at any location in time and space. Additionally, we introduce an adaptive collocation strategy in space to prioritize sampling from the most important regions. GraphSplineNets improve the accuracy-speedup tradeoff in forecasting various dynamical systems with increasing complexity, including the heat equation, damped wave propagation, Navier-Stokes equations, and real-world ocean currents in both regular and irregular domains.

We study the problem of optimizing biological sequences, e.g., proteins, DNA, and RNA, to maximize a black-box score function that is only evaluated in an offline dataset. We propose a novel solution, bootstrapped training of score-conditioned generator (BootGen) algorithm. Our algorithm repeats a two-stage process. In the first stage, our algorithm trains the biological sequence generator with rank-based weights to enhance the accuracy of sequence generation based on high scores. The subsequent stage involves bootstrapping, which augments the training dataset with self-generated data labeled by a proxy score function. Our key idea is to align the score-based generation with a proxy score function, which distills the knowledge of the proxy score function to the generator. After training, we aggregate samples from multiple bootstrapped generators and proxies to produce a diverse design. Extensive experiments show that our method outperforms competitive baselines on biological sequential design tasks. We provide reproducible source code: https://github.com/kaist-silab/bootgen.

Based on the Fourier duality between a stationary kernel and its spectral density, modeling the spectral density using a Gaussian mixture density enables one to construct a flexible kernel, known as a Spectral Mixture kernel, that can model any stationary kernel. However, despite its expressive power, training this kernel is typically difficult because scalability and overfitting issues often arise due to a large number of training parameters. To resolve these issues, we propose an approximate inference method for estimating the Spectral mixture kernel hyperparameters. Specifically, we approximate this kernel by using the finite random spectral points based on Random Fourier Feature and optimize the parameters for the distribution of spectral points by sampling-based variational inference. To improve this inference procedure, we analyze the training loss and propose two special methods: a sampling method of spectral points to reduce the error of the approximate kernel in training, and an approximate natural gradient to accelerate the convergence of parameter inference.

This paper explores the possibility of near-optimally solving multi-agent, multi-task NP-hard planning problems with time-dependent rewards using a learning-based algorithm. In particular, we consider a class of robot/machine scheduling problems called the multi-robot reward collection problem (MRRC). Such MRRC problems well model ride-sharing, pickup-and-delivery, and a variety of related problems. In representing the MRRC problem as a sequential decision-making problem, we observe that each state can be represented as an extension of probabilistic graphical models (PGMs), which we refer to as random PGMs. We then develop a mean-field inference method for random PGMs. We then propose (1) an order-transferable Q-function estimator and (2) an order-transferability-enabled auction to select a joint assignment in polynomial-time. These result in a reinforcement learning framework with at least $1-1/e$ optimality. Experimental results on solving MRRC problems highlight the near-optimality and transferability of the proposed methods. We also consider identical parallel machine scheduling problems (IPMS) and minimax multiple traveling salesman problems (minimax-mTSP).

Deep reinforcement learning (DRL)-based combinatorial optimization (CO) methods (i.e., DRL-NCO) have shown significant merit over the conventional CO solvers as DRL-NCO is capable of learning CO solvers less relying on problem-specific expert domain knowledge (heuristic method) and supervised labeled data (supervised learning method). This paper presents a novel training scheme, Sym-NCO, which is a regularizer-based training scheme that leverages universal symmetricities in various CO problems and solutions. Leveraging symmetricities such as rotational and reflectional invariance can greatly improve the generalization capability of DRL-NCO because it allows the learned solver to exploit the commonly shared symmetricities in the same CO problem class. Our experimental results verify that our Sym-NCO greatly improves the performance of DRL-NCO methods in four CO tasks, including the traveling salesman problem (TSP), capacitated vehicle routing problem (CVRP), prize collecting TSP (PCTSP), and orienteering problem (OP), without utilizing problem-specific expert domain knowledge. Remarkably, Sym-NCO outperformed not only the existing DRL-NCO methods but also a competitive conventional solver, the iterative local search (ILS), in PCTSP at 240$\times$ faster speed. Our source code is available at https://github.com/alstn12088/Sym-NCO.

Spectral analysis provides one of the most effective paradigms for information-preserving dimensionality reduction, as simple descriptions of naturally occurring signals are often obtained via few terms of periodic basis functions. In this work, we study deep neural networks designed to harness the structure in frequency domain for efficient learning of long-range correlations in space or time: frequency-domain models (FDMs). Existing FDMs are based on complex-valued transforms i.e. Fourier Transforms (FT), and layers that perform computation on the spectrum and input data separately. This design introduces considerable computational overhead: for each layer, a forward and inverse FT. Instead, this work introduces a blueprint for frequency domain learning through a single transform: transform once (T1). To enable efficient, direct learning in the frequency domain we derive a variance preserving weight initialization scheme and investigate methods for frequency selection in reduced-order FDMs. Our results noticeably streamline the design process of FDMs, pruning redundant transforms, and leading to speedups of 3x to 10x that increase with data resolution and model size. We perform extensive experiments on learning the solution operator of spatio-temporal dynamics, including incompressible Navier-Stokes, turbulent flows around airfoils and high-resolution video of smoke. T1 models improve on the test performance of FDMs while requiring significantly less computation (5 hours instead of 32 for our large-scale experiment), with over 20% reduction in predictive error across tasks.

In this study, we propose an adversarially-trained phase-consistent network (APCNet), which is a semi-supervised signal classification approach. The proposed classification model is trained with datasets that contain a small fraction of labeled output so as to design (1) an effective representation of the input time series (vibration signal) to extract important factors for the model to discriminate between different bearing conditions, and (2) a latent representation for the data to reflect the true data distribution precisely. To achieve these goals, APCNet suggests three novelties: the vibration-specific encoder, the phase-consistency regularization, and the adversarially-trained latent distribution alignment of the labeled and unlabeled distributions. We conduct experiments on two public bearing datasets and one public motor operating dataset to evaluate the performance of APCNet. We interpret the model's capabilities with different data label ratios and latent distribution analysis. The results show that APCNet performs well on datasets with small labeled to unlabeled data ratio. Also, we show that APCNet achieves our objectives of capturing important vibration signals features and modeling the true data distribution effectively.

Recently, deep reinforcement learning (DRL) frameworks have shown potential for solving NP-hard routing problems such as the traveling salesman problem (TSP) without problem-specific expert knowledge. Although DRL can be used to solve complex problems, DRL frameworks still struggle to compete with state-of-the-art heuristics showing a substantial performance gap. This paper proposes a novel hierarchical problem-solving strategy, termed learning collaborative policies (LCP), which can effectively find the near-optimum solution using two iterative DRL policies: the seeder and reviser. The seeder generates as diversified candidate solutions as possible (seeds) while being dedicated to exploring over the full combinatorial action space (i.e., sequence of assignment action). To this end, we train the seeder's policy using a simple yet effective entropy regularization reward to encourage the seeder to find diverse solutions. On the other hand, the reviser modifies each candidate solution generated by the seeder; it partitions the full trajectory into sub-tours and simultaneously revises each sub-tour to minimize its traveling distance. Thus, the reviser is trained to improve the candidate solution's quality, focusing on the reduced solution space (which is beneficial for exploitation). Extensive experiments demonstrate that the proposed two-policies collaboration scheme improves over single-policy DRL framework on various NP-hard routing problems, including TSP, prize collecting TSP (PCTSP), and capacitated vehicle routing problem (CVRP).

We detail a novel class of implicit neural models. Leveraging time-parallel methods for differential equations, Multiple Shooting Layers (MSLs) seek solutions of initial value problems via parallelizable root-finding algorithms. MSLs broadly serve as drop-in replacements for neural ordinary differential equations (Neural ODEs) with improved efficiency in number of function evaluations (NFEs) and wall-clock inference time. We develop the algorithmic framework of MSLs, analyzing the different choices of solution methods from a theoretical and computational perspective. MSLs are showcased in long horizon optimal control of ODEs and PDEs and as latent models for sequence generation. Finally, we investigate the speedups obtained through application of MSL inference in neural controlled differential equations (Neural CDEs) for time series classification of medical data.

Effective control and prediction of dynamical systems require appropriate handling of continuous-time and discrete, event-triggered processes. Stochastic hybrid systems (SHSs), common across engineering domains, provide a formalism for dynamical systems subject to discrete, possibly stochastic, state jumps and multi-modal continuous-time flows. Despite the versatility and importance of SHSs across applications, a general procedure for the explicit learning of both discrete events and multi-mode continuous dynamics remains an open problem. This work introduces Neural Hybrid Automata (NHAs), a recipe for learning SHS dynamics without a priori knowledge on the number, mode parameters, and inter-modal transition dynamics. NHAs provide a systematic inference method based on normalizing flows, neural differential equations, and self-supervision. We showcase NHAs on several tasks, including mode recovery and flow learning in systems with stochastic transitions, and end-to-end learning of hierarchical robot controllers.

Most previous studies on multi-agent reinforcement learning focus on deriving decentralized and cooperative policies to maximize a common reward and rarely consider the transferability of trained policies to new tasks. This prevents such policies from being applied to more complex multi-agent tasks. To resolve these limitations, we propose a model that conducts both representation learning for multiple agents using hierarchical graph attention network and policy learning using multi-agent actor-critic. The hierarchical graph attention network is specially designed to model the hierarchical relationships among multiple agents that either cooperate or compete with each other to derive more advanced strategic policies. Two attention networks, the inter-agent and inter-group attention layers, are used to effectively model individual and group level interactions, respectively. The two attention networks have been proven to facilitate the transfer of learned policies to new tasks with different agent compositions and allow one to interpret the learned strategies. Empirically, we demonstrate that the proposed model outperforms existing methods in several mixed cooperative and competitive tasks.

In this study, we present a hypergraph convolutional recurrent neural network (HGC-RNN), which is a prediction model for structured time-series sensor network data. Representing sensor networks in a graph structure is useful for expressing structural relationships among sensors. Conventional graph structure, however, has a limitation on representing complex structure in real world application, such as shared connections among multiple nodes. We use a hypergraph, which is capable of modeling complicated structures, for structural representation. HGC-RNN performs a hypergraph convolution operation on the input data represented in the hypergraph to extract hidden representations of the input, while considering the structural dependency of the data. HGC-RNN employs a recurrent neural network structure to learn temporal dependency from the data sequence. We conduct experiments to forecast taxi demand in NYC, traffic flow in the overhead hoist transfer system, and gas pressure in a gas regulator. We compare the performance of our method with those of other existing methods, and the result shows that HGC-RNN has strengths over baseline models.

Continuous deep learning architectures have recently re-emerged as Neural Ordinary Differential Equations (Neural ODEs). This infinite-depth approach theoretically bridges the gap between deep learning and dynamical systems, offering a novel perspective. However, deciphering the inner working of these models is still an open challenge, as most applications apply them as generic black-box modules. In this work we ``open the box'', further developing the continuous-depth formulation with the aim of clarifying the influence of several design choices on the underlying dynamics.

The infinite-depth paradigm pioneered by Neural ODEs has launched a renaissance in the search for novel dynamical system-inspired deep learning primitives; however, their utilization in problems of non-trivial size has often proved impossible due to poor computational scalability. This work paves the way for scalable Neural ODEs with time-to-prediction comparable to traditional discrete networks. We introduce hypersolvers, neural networks designed to solve ODEs with low overhead and theoretical guarantees on accuracy. The synergistic combination of hypersolvers and Neural ODEs allows for cheap inference and unlocks a new frontier for practical application of continuous-depth models. Experimental evaluations on standard benchmarks, such as sampling for continuous normalizing flows, reveal consistent pareto efficiency over classical numerical methods.

This paper comprises a description of a data-driven approach to the real-time monitoring of a physical system. Specifically, a hierarchical anomaly detection algorithm that can identify both instantaneous pointwise anomalies and gradual trajectory anomalies is proposed. To detect anomalies, we first construct a multioutput Gaussian process regression (MOGPR) model that can predict, probabilistically, the outputs of the target system. Using the constructed prediction model, we then propose the statistical decision-making strategies to determine the abnormal operations of the target system by comparing its measured and the predicted responses. For pointwise anomaly detection, we regard a single measurement as abnormal if the difference between the measurement and the prediction exceeds the threshold based on an extreme value theory. For the trajectory anomaly detection, we consider a sequence of measurements abnormal if the Mahalanobis distance between the measured and predicted trajectories is highly improbable. The proposed monitoring strategy does both the pointwise and the trajectory anomaly detection in a single framework. The proposed strategy was applied to detecting abnormal operations of gas regulators. Validating with the actual gas regulator data demonstrated that it could identify the anomalies robustly and accurately. Note to Practitioners-This paper was motivated by the problem of a need to detect abnormal operations in industrial gas regulators, but it also applies to other industrial mechanical systems. Existing approaches to identifying abnormal operations in such systems are premised on rule-based methods that issue alarms if the measured response signal from a target system exceeds a certain fixed threshold. We propose an anomaly detection algorithm, constructed on the basis of a multioutput Gaussian process regression that takes into account the correlations of multiple-target time series data. In particular, the proposed method makes it possible to detect both pointwise anomalies (outlier) and trajectory anomalies (i.e., a systemic deviation between the measured and predicted target responses) in a single framework while adapting its prediction model with newly observed data. The proposed method was tested using data recorded from currently operating gas regulators. It was found that this method enabled one to detect the abnormal operations of the target system reliably. In the future research, we will accelerate the learning and adapt the speed of the target system so that the method can be quickly run in a real-time system.

This paper proposes a deep reinforcement learning approach in order to optimize a sequence of tasks efficiently with the aid of image processing techniques used in computer vision. The proposed algorithm can be employed to solve the traveling salesman problem (TSP), a combinatorial optimization problem that determines the optimum trajectory of city visits so that the total traveling distance is minimized. The proposed algorithm accepts a set of images as an input, and outputs the priority over alternative tasks (or sites to visit) that should be conducted at the next time step. The proposed method applies a stacked convolutional network layer to effectively process and extract the meaningful features and uses a fully convolutional network to map the processed features to the output tasks without losing the local connectivity in the input images. The proposed algorithm has been employed to optimize the excavation schedule of a single digger for completing a 20 by 20 grid world, which is equivalent to the TSP problem with a node size of 400. The simulation results showed that the proposed method can achieve an effective schedule with optimality comparable to state of the art algorithms.

The use of data-driven predictive models is becoming increasingly popular in engineering and manufacturing sectors. This paper discusses the deployment of Gaussian Process Regression (GPR) predictive models for smart manufacturing. A scoring engine is developed based on the Predictive Model Markup Language (PMML) standard to illustrate the portability of predictive models among different statistical tools and different platforms. Specifically, we evaluate the tradeoffs between embedding GPR-based predictive models on a physical device and executing the predictive models on a managed cloud platform like the Google Compute Engine. We compare the performance of the two deployment strategies with two predictive models, namely an energy consumption model and a milling tool condition model, that are built with data from a Mori Seiki CNC milling machine. We describe how the response time of the two deployment strategies is related to the network latency and computational speed of the scoring machine hardware. It is shown that the time required to calculate model predictions is a significant factor in the overall response time of the embedded scoring engine. We demonstrate that the scoring engine on the cloud platform can achieve a lower response time and higher prediction rate than the microcomputer, due to the superior computational performance of the cloud-based hardware.

This paper describes a real-time data collection framework and an adaptive machining learning method for constructing a real-time energy prediction model for a machine tool. To effectively establish the energy consumption pattern of a machine tool over time, the energy prediction model is continuously updated with new measurement data to account for time-varying effects of the machine tool, such as tool wear and machine tool deterioration. In this work, a real-time data collection and processing framework is developed to retrieve raw data from a milling machine tool and its sensors and convert them into relevant input features. The extracted input features are then used to construct the energy prediction model using Gaussian Process (GP) regression. To update the GP regression model with real-time streaming data, we investigate the use of sparse representation of the covariance matrix to reduce the computational and storage demands of the GP regression. We compare computational efficiency of sparse GP to that of full GP regression model and show the effectiveness of the sparse GP regression model for tracking the variation in the energy consumption pattern of the target machine.

Recent advances in machine automation and sensing technology offer new opportunities for continuous condition monitoring of an operating machine. This paper describes an intelligent machine monitoring framework that integrates and utilizes data collection, management, and analytics to derive an adaptive predictive model for the energy usage of a milling machine. This model is designed using a Gaussian Process (GP) regression algorithm, which is a flexible regression method that also provides an uncertainty estimate. To improve computational efficiency, we propose a Collective Gaussian Process (CGP) in which the overall energy prediction is made by constructing local GP models weighted by probability distribution functions obtained using the Gaussian Mixture Model (GMM) technique. Finally, we demonstrate the ability of the proposed monitoring framework to construct an energy prediction model to predict the energy used to machine a part.