Planning is essential for artificial intelligence systems to look ahead and proactively determine a course of actions to reach objectives in the virtual and real world. Recent work on large language models (LLMs) sheds light on their planning capability in various tasks. However, it remains unclear what signals in the context influence the model performance. In this work, we explore how to improve the model planning capability through in-context learning (ICL), specifically, what signals can help select the exemplars. Through extensive experiments, we observe that commonly used problem similarity may result in false positives with drastically different plans, which can mislead the model. In response, we propose to sample and filter exemplars leveraging plan side action sequence similarity (AS). We propose GRASE-DC: a two-stage pipeline that first re-samples high AS exemplars and then curates the selected exemplars with dynamic clustering on AS to achieve a balance of relevance and diversity. Our experimental result confirms that GRASE-DC achieves significant performance improvement on various planning tasks (up to ~11-40 point absolute accuracy improvement with 27.3% fewer exemplars needed on average). With GRASE-DC* + VAL, where we iteratively apply GRASE-DC with a validator, we are able to even boost the performance by 18.9% more. Extensive analysis validates the consistent performance improvement of GRASE-DC with various backbone LLMs and on both classical planning and natural language planning benchmarks. GRASE-DC can further boost the planning accuracy by ~24 absolute points on harder problems using simpler problems as exemplars over a random baseline. This demonstrates its ability to generalize to out-of-distribution problems.

Selecting the best items in a dataset is a common task in data exploration. However, the concept of “best” lies in the eyes of the beholder: Different users may consider different attributes more important and, hence, arrive at different rankings. Nevertheless, one can remove “dominated” items and create a “representative” subset of the data, comprising the “best items” in it. A Pareto-optimal representative is guaranteed to contain the best item of each possible ranking, but it can be a large portion of data. A much smaller representative can be found if we relax the requirement of including the best item for each user and instead just limit the users’ “regret.” Existing work defines regret as the loss in score by limiting consideration to the representative instead of the full dataset, for any chosen ranking function. However, the score is often not a meaningful number, and users may not understand its absolute value. Sometimes small ranges in score can include large fractions of the dataset. In contrast, users do understand the notion of rank ordering. Therefore, we consider items’ positions in the ranked list in defining the regret and propose the rank-regret representative as the minimal subset of the data containing at least one of the top- k of any possible ranking function. This problem is polynomial time solvable in two-dimensional space but is NP-hard on three or more dimensions. We design a suite of algorithms to fulfill different purposes, such as whether relaxation is permitted on k , the result size, or both, whether a distribution is known, whether theoretical guarantees or practical efficiency is important, and so on. Experiments on real datasets demonstrate that we can efficiently find small subsets with small rank-regrets.

Constrained random verification in industrial settings involves parameterized tests. The parameters used to control the test stimuli generation are typically set when the test is first written, and seldomly varied later on in nightly regressions. In this work, we formulate test parameter configuration as a blackbox optimization problem and introduce Smart Regression Planner (SRP), an approach that automatically configures the tunable test parameters to better explore the input space and accelerate convergence towards coverage. The optimizer in SRP can drive the parameters update with a Bayesian optimization technique that uses coverage from nightly regressions as feedback. Our evaluation on open-source as well as larger industrial designs demonstrates that SRP leads to up to 9.84% higher average coverage over 100 nights than the human baseline. Importantly, it converges to coverage milestones up to 20 × faster than the human baseline. With high-level test parameter optimization, we introduce a problem space and an opportunity to achieve categorically higher coverage in industrial settings with very low overhead. Furthermore, through two practical use cases, we show that employing multi-objective optimization and transfer learning can further accelerate the verification process. to structure and flexible enough to adapt to changes in an evolving design. While random search relies purely on exploration, Bayesian optimization exploits learning through feedback. We also investigate use cases of (1) simultaneously minimizing runtime and maximizing coverage using multi-objective Bayesian optimization, and (2) transfer learning, or the ability to transfer learned heuristics from one set of parameters to another through design evolution. practical impact, at an abstraction level where scalability is a natural byproduct. We have shown that employing Bayesian optimization to adjust parameters in a testbench allows us to improve coverage in real industrial settings. The improvement comes at a very low engineering overhead: the system can be assembled from standard commercial RTL simulators and off-the-shelf blackbox optimizers, and involves no effort from the verification engineers once set up. Furthermore, we show that through multi-objective optimization, it is possible to improve both test run-time and coverage (leading to more efficient use of simulator licenses and compute), and through transfer learning, it can effectively handle continuously evolving designs and testbenches (as is a requirement in practice).

Signal reconstruction problem (SRP) is an important optimization problem where the objective is to identify a solution to an underdetermined system of linear equations that is closest to a given prior. It has a substantial number of applications in diverse areas, such as network traffic engineering, medical image reconstruction, acoustics, astronomy, and many more. Unfortunately, most of the common approaches for solving SRP do not scale to large problem sizes. We propose a novel and scalable algorithm for solving this critical problem. Specifically, we make four major contributions. First, we propose a dual formulation of the problem and develop the DIRECT algorithm that is significantly more efficient than the state of the art. Second, we show how adapting database techniques developed for scalable similarity joins provides a substantial speedup over DIRECT. Third, we describe several practical techniques that allow our algorithm to scale-on a single machine-to settings that are orders of magnitude larger than previously studied. Finally, we use the database techniques of materialization and reuse to extend our result to dynamic settings where the input to the SRP changes. Extensive experiments on real-world and synthetic data confirm the efficiency, effectiveness, and scalability of our proposal.

Constrained random verification (CRV) in industrial settings involves manual parameterized test generation, a costly and inefficient process. We formulate test parameter configuration as a blackbox optimization problem and we introduce Smart Regression Planner (SRP), an approach that automatically configures the tunable test parameters to better explore the input space and accelerate convergence towards coverage. The optimizer in SRP can drive the parameters update with twomethods: a light-weight random search, and a Bayesian optimization technique that uses coverage from nightly regressions as feedback. Our experimental evaluation on open-source as well as larger industrial designs demonstrates that frequent perturbation and optimization of test parameters leads to higher coverage than the human baseline. Importantly, it converges to coverage milestones significantly faster than the human baseline. With high-level test parameter optimization, we introduce a problem space and an opportunity to achieve categorically higher coverage in industrial settings with very low overhead.

Chip floorplanning is the engineering task of designing the physical layout of a computer chip. Despite five decades of research 1 , chip floorplanning has defied automation, requiring months of intense effort by physical design engineers to produce manufacturable layouts. Here we present a deep reinforcement learning approach to chip floorplanning. In under six hours, our method automatically generates chip floorplans that are superior or comparable to those produced by humans in all key metrics, including power consumption, performance and chip area. To achieve this, we pose chip floorplanning as a reinforcement learning problem, and develop an edge-based graph convolutional neural network architecture capable of learning rich and transferable representations of the chip. As a result, our method utilizes past experience to become better and faster at solving new instances of the problem, allowing chip design to be performed by artificial agents with more experience than any human designer. Our method was used to design the next generation of Google’s artificial intelligence (AI) accelerators, and has the potential to save thousands of hours of human effort for each new generation. Finally, we believe that more powerful AI-designed hardware will fuel advances in AI, creating a symbiotic relationship between the two fields.

Learning graph generative models is a challenging task for deep learning and has wide applicability to a range of domains like chemistry, biology and social science. However current deep neural methods suffer from limited scalability: for a graph with n nodes and m edges, existing deep neural methods require Omega(n(2)) complexity by building up the adjacency matrix. On the other hand, many real world graphs are actually sparse in the sense that m << n(2) . Based on this, we develop a novel autoregressive model, named BiGG, that utilizes this sparsity to avoid generating the full adjacency matrix, and importantly reduces the graph generation time complexity to O((n m) log n). Furthermore, during training this autoregressive model can be parallelized with O(log n) synchronization stages, which makes it much more efficient than other autoregressive models that require Omega(n). Experiments on several benchmarks show that the proposed approach not only scales to orders of magnitude larger graphs than previously possible with deep autoregressive graph generative models, but also yields better graph generation quality.

We introduce a novel end-to-end approach for learning to cluster in the absence of labeled examples. Our clustering objective is based on optimizing normalized cuts, a criterion which measures both intra-cluster similarity as well as inter-cluster dissimilarity. We define a differentiable loss function equivalent to the expected normalized cuts. Unlike much of the work in unsupervised deep learning, our trained model directly outputs final cluster assignments, rather than embeddings that need further processing to be usable. Our approach generalizes to unseen datasets across a wide variety of domains, including text, and image. Specifically, we achieve state-of-the-art results on popular unsupervised clustering benchmarks (e.g., MNIST, Reuters, CIFAR-10, and CIFAR-100), outperforming the strongest baselines by up to 10.9%. Our generalization results are superior (by up to 21.9%) to the recent top-performing clustering approach with the ability to generalize.

Reconstructing a high dimensional unknown signal, using lower dimensional observations is a challenging problem, known as signal reconstruction problem (SRP), with diverse applications including network traffic engineering, medical image reconstruction, and astronomy. Recently the database community has shown significant advancements in solving the SRP problem efficiently, effectively, and in scale by leveraging database techniques such as similarity joins. In this demo, we demonstrate Orca-SR that highlights the benefits of signal reconstruction in scale by demonstrating real-time network traffic flow analysis on large networks that were not possible before. Orca-SR is a web application that enables a user to generate network flow and load the network for interactive analysis of the impact of different traffic patterns on signal reconstruction.

The signal reconstruction problem (SRP) is an important optimization problem where the objective is to identify a solution to an under-determined system of linear equations AX = b that is closest to a given prior. It has a substantial number of applications in diverse areas including network traffic engineering, medical image reconstruction, acoustics, astronomy and many more. Most common approaches for solving SRP do not scale to large problem sizes. In this paper, we propose a dual formulation of this problem and show how adapting database techniques developed for scalable similarity joins provides a significant speedup when the A matrix is sparse and binary. Extensive experiments on real-world and synthetic data show that our approach produces a significant speedup of up to 20x over competing approaches.

Peer to peer marketplaces such as AirBnB enable transactional exchange of services directly between people. In such platforms, those providing a service (hosts in AirBnB) are faced with various choices. For example in AirBnB, although some amenities in a property (attributes of the property) are fixed, others are relatively flexible and can be provided without significant effort. Providing an amenity is usually associated with a cost. Naturally different sets of amenities may have a different "gains" for a host. Consequently, given a limited budget, deciding which amenities (attributes) to offer is challenging. In this paper, we formally introduce and define the problem of Gain Maximization over Flexible Attributes (GMFA). We first prove that the problem is NP-hard and show that identifying an approximate algorithm with a constant approximate ratio is unlikely. We then provide a practically efficient exact algorithm to the GMFA problem for the general class of monotonic gain functions, which quantify the benefit of sets of attributes. As the next part of our contribution, we focus on the design of a practical gain function for GMFA. We introduce the notion of frequent-item based count (FBC), which utilizes the existing tuples in the database to define the notion of gain, and propose an efficient algorithm for computing it. We present the results of a comprehensive experimental evaluation of the proposed techniques on real dataset from AirBnB and demonstrate the practical relevance and utility of our proposal.

Signal reconstruction problem (SRP) is an important optimization problem where the objective is to identify a solution to an underdetermined system of linear equations that is closest to a given prior. It has a substantial number of applications in diverse areas including network traffic engineering, medical image reconstruction, acoustics, astronomy and many more. Most common approaches for SRP do not scale to large problem sizes. In this paper, we propose multiple optimization steps, developing scalable algorithms for the problem. We first propose a dual formulation of the problem and develop the Direct algorithm that is significantly more efficient than the state of the art. Second, we show how adapting database techniques developed for scalable similarity joins provides a significant speedup over Direct, scaling our proposal up to large-scale settings. Third, we describe a number of practical techniques that allow our algorithm to scale to settings of size in the order of a million by a billion. We also adapt our proposal to identify the top-k components of the solved system of linear equations. Finally, we consider the dynamic setting where the inputs to the linear system change and propose efficient algorithms inspired by the database techniques of materialization and reuse. Extensive experiments on real-world and synthetic data confirm the efficiency, effectiveness and scalability of our proposal.

Query optimizers depend on selectivity estimates of query predicates to produce a good execution plan. When a query contains multiple predicates, today's optimizers use a variety of assumptions, such as independence between predicates, to estimate selectivity. While such techniques have the benefit of fast estimation and small memory footprint, they often incur large selectivity estimation errors. In this work, we reconsider selectivity estimation as a regression problem. We explore application of neural networks and tree-based ensembles to the important problem of selectivity estimation of multi-dimensional range predicates. While their straightforward application does not outperform even simple baselines, we propose two simple yet effective design choices, i.e., regression label transformation and feature engineering, motivated by the selectivity estimation context. Through extensive empirical evaluation across a variety of datasets, we show that the proposed models deliver both highly accurate estimates as well as fast estimation.

Graph partitioning is the problem of dividing the nodes of a graph into balanced partitions while minimizing the edge cut across the partitions. Due to its combinatorial nature, many approximate solutions have been developed, including variants of multi-level methods and spectral clustering. We propose GAP, a Generalizable Approximate Partitioning framework that takes a deep learning approach to graph partitioning. We define a differentiable loss function that represents the partitioning objective and use backpropagation to optimize the network parameters. Unlike baselines that redo the optimization per graph, GAP is capable of generalization, allowing us to train models that produce performant partitions at inference time, even on unseen graphs. Furthermore, because we learn the representation of the graph while jointly optimizing for the partitioning loss function, GAP can be easily tuned for a variety of graph structures. We evaluate the performance of GAP on graphs of varying sizes and structures, including graphs of widely used machine learning models (e.g., ResNet, VGG, and Inception-V3), scale-free graphs, and random graphs. We show that GAP achieves competitive partitions while being up to 100 times faster than the baseline and generalizes to unseen graphs.

Selecting the best items in a dataset is a common task in data exploration. However, the concept of "best'' lies in the eyes of the beholder: different users may consider different attributes more important, and hence arrive at different rankings. Nevertheless, one can remove "dominated'' items and create a "representative'' subset of the data, comprising the "best items'' in it. A Pareto-optimal representative is guaranteed to contain the best item of each possible ranking, but it can be a large portion of data. A much smaller representative can be found if we relax the requirement to include the best item for each user, and instead just limit the users' "regret''. Existing work defines regret as the loss in score by limiting consideration to the representative instead of the full data set, for any chosen ranking function. However, the score is often not a meaningful number and users may not understand its absolute value. Sometimes small ranges in score can include large fractions of the data set. In contrast, users do understand the notion of rank ordering. Therefore, we consider the position of the items in the ranked list for defining the regret and propose the \em rank-regret representative as the minimal subset of the data containing at least one of the top-k of any possible ranking function. This problem is NP-complete. We use a geometric interpretation of items to bound their ranks on ranges of functions and to utilize combinatorial geometry notions for developing effective and efficient approximation algorithms for the problem. Experiments on real datasets demonstrate that we can efficiently find small subsets with small rank-regrets.

Efficiently estimating the inclusion coefficient - the fraction of values of one column that are contained in another column - is useful for tasks such as data profiling and foreign-key detection. We present a new estimator, BML, for inclusion coefficient based on Hyperloglog sketches that results in significantly lower error compared to the state-of-the art approach that uses Bottom-k sketches. We evaluate the error of the BML estimator using experiments on industry benchmarks such as TPC-H and TPC-DS, and several real-world databases. As an independent contribution, we show how Hyperloglog sketches can be maintained incrementally with data deletions using only a constant amount of additional memory.

Signal reconstruction problem (SRP) is an important optimization problem where the objective is to identify a solution to an underdetermined system of linear equations that is closest to a given prior. It has a substantial number of applications in diverse areas including network traffic engineering, medical image reconstruction, acoustics, astronomy and many more. Most common approaches for SRP do not scale to large problem sizes. In this paper, we propose a dual formulation of this problem and show how adapting database techniques developed for scalable similarity joins provides a significant speedup. Extensive experiments on real-world and synthetic data show that our approach produces a significant speedup of up to 20x over competing approaches.

A large number of web databases are hidden behind form-based interfaces (consisting of interface components such as textboxes, drop-down boxes, etc) for users to enter their desired values in a search query. Form-based interfaces are not always easy to use and error-prone on mobile devices, mainly because of limitations such as smaller screen, trickier text input, etc. We build MobiFace which is a faceted search system over a hidden database for mobile users. The goal of our system is to help a user to find her tuple of interest with as fewer questions as possible. MobiFace dynamically suggests facets for drilling down into the hidden web database such that the navigation cost is minimized. We run the experiment over real IMDB dataset on a mobile device.

Finding the maxima of a database based on a user preference, especially when the ranking function is a linear combination of the attributes, has been the subject of recent research. A critical observation is that the em convex hull is the subset of tuples that can be used to find the maxima of any linear function. However, in real world applications the convex hull can be a significant portion of the database, and thus its performance is greatly reduced. Thus, computing a subset limited to $r$ tuples that minimizes the regret ratio (a measure of the user's dissatisfaction with the result from the limited set versus the one from the entire database) is of interest. In this paper, we make several fundamental theoretical as well as practical advances in developing such a compact set. In the case of two dimensional databases, we develop an optimal linearithmic time algorithm by leveraging the ordering of skyline tuples. In the case of higher dimensions, the problem is known to be NPcomplete. As one of our main results of this paper, we develop an approximation algorithm that runs in linearithmic time and guarantees a regret ratio, within any arbitrarily small user-controllable distance from the optimal regret ratio. The comprehensive set of experiments on both synthetic and publicly available real datasets confirm the efficiency, quality of output, and scalability of our proposed algorithms.