Datasets:

ai-conferences
/

ICML2023

paper_id stringlengths 10 10	title stringlengths 10 168	paper_url stringlengths 42 42	authors listlengths 1 42	type stringclasses 2 values	primary_area stringclasses 0 values	abstract large_stringlengths 417 2.06k	keywords listlengths 0 0	TL;DR large_stringclasses 0 values	submission_number int64 6 6.84k	arxiv_id stringlengths 10 10 ⌀	arxiv_id_source stringclasses 2 values
PwsvGnamYD	Bayesian Estimation of Differential Privacy	https://openreview.net/forum?id=PwsvGnamYD	[ "Santiago Zanella-Beguelin", "Lukas Wutschitz", "Shruti Tople", "Ahmed Salem", "Victor Rühle", "Andrew Paverd", "Mohammad Naseri", "Boris Köpf", "Daniel Jones" ]	Poster	null	Algorithms such as Differentially Private SGD enable training machine learning models with formal privacy guarantees. However, because these guarantees hold with respect to unrealistic adversaries, the protection afforded against practical attacks is typically much better. An emerging strand of work empirically estimat...	[]	null	6,837	2206.05199	title_snapshot
USiX9gmGRx	Adaptive Estimation of Graphical Models under Total Positivity	https://openreview.net/forum?id=USiX9gmGRx	[ "Jiaxi Ying", "José Vinícius De Miranda Cardoso", "Daniel P. Palomar" ]	Poster	null	We consider the problem of estimating (diagonally dominant) M-matrices as precision matrices in Gaussian graphical models. Such models have shown interesting properties, e.g., the maximum likelihood estimator exists with as little as two observations in the case of M-matrices, and exists even with one observation in th...	[]	null	6,836	2210.15471	title_snapshot
elL6uw9qOX	GEAR: A GPU-Centric Experience Replay System for Large Reinforcement Learning Models	https://openreview.net/forum?id=elL6uw9qOX	[ "Hanjing Wang", "Man-Kit Sit", "Congjie He", "Ying Wen", "Weinan Zhang", "Jun Wang", "Yaodong Yang", "Luo Mai" ]	Poster	null	This paper introduces a distributed, GPU-centric experience replay system, GEAR, designed to perform scalable reinforcement learning (RL) with large sequence models (such as transformers). With such models, existing systems such as Reverb face considerable bottlenecks in memory, computation, and communication. GEAR, ho...	[]	null	6,829	2310.05205	title_snapshot
jOLIFanYnt	Disentangled Multi-Fidelity Deep Bayesian Active Learning	https://openreview.net/forum?id=jOLIFanYnt	[ "Dongxia Wu", "Ruijia Niu", "Matteo Chinazzi", "Yian Ma", "Rose Yu" ]	Poster	null	To balance quality and cost, various domain areas of science and engineering run simulations at multiple levels of sophistication. Multi-fidelity active learning aims to learn a direct mapping from input parameters to simulation outputs at the highest fidelity by actively acquiring data from multiple fidelity levels. H...	[]	null	6,826	2305.04392	title_snapshot
ikE60aXe8M	Understand and Modularize Generator Optimization in ELECTRA-style Pretraining	https://openreview.net/forum?id=ikE60aXe8M	[ "Chengyu Dong", "Liyuan Liu", "Hao Cheng", "Jingbo Shang", "Jianfeng Gao", "Xiaodong Liu" ]	Poster	null	Despite the effectiveness of ELECTRA-style pre-training, their performance is dependent on the careful selection of the model size for the auxiliary generator, leading to high trial-and-error costs. In this paper, we present the first systematic study of this problem. Our theoretical investigation highlights the import...	[]	null	6,816	null	null
cHhGmXDiHp	NeRFool: Uncovering the Vulnerability of Generalizable Neural Radiance Fields against Adversarial Perturbations	https://openreview.net/forum?id=cHhGmXDiHp	[ "Yonggan Fu", "Ye Yuan", "Souvik Kundu", "Shang Wu", "Shunyao Zhang", "Celine Lin" ]	Poster	null	Generalizable Neural Radiance Fields (GNeRF) are one of the most promising real-world solutions for novel view synthesis, thanks to their cross-scene generalization capability and thus the possibility of instant rendering on new scenes. While adversarial robustness is essential for real-world applications, little study...	[]	null	6,810	2306.06359	title_snapshot
MI5YpKX84O	Provably Efficient Representation Learning with Tractable Planning in Low-Rank POMDP	https://openreview.net/forum?id=MI5YpKX84O	[ "Jiacheng Guo", "Zihao Li", "Huazheng Wang", "Mengdi Wang", "Zhuoran Yang", "Xuezhou Zhang" ]	Poster	null	In this paper, we study representation learning in partially observable Markov Decision Processes (POMDPs), where the agent learns a decoder function that maps a series of high-dimensional raw observations to a compact representation and uses it for more efficient exploration and planning. We focus our attention on the...	[]	null	6,793	2306.12356	title_snapshot
H21qm4xyk9	Taming graph kernels with random features	https://openreview.net/forum?id=H21qm4xyk9	[ "Krzysztof Marcin Choromanski" ]	Oral	null	We introduce in this paper the mechanism of graph random features (GRFs). GRFs can be used to construct unbiased randomized estimators of several important kernels defined on graphs' nodes, in particular the regularized Laplacian kernel. As regular RFs for non-graph kernels, they provide means to scale up kernel method...	[]	null	6,792	2305.00156	title_snapshot
EMN99LtfYA	Block Subsampled Randomized Hadamard Transform for Nyström Approximation on Distributed Architectures	https://openreview.net/forum?id=EMN99LtfYA	[ "Oleg Balabanov", "Matthias Beaupère", "Laura Grigori", "Victor Lederer" ]	Poster	null	This article introduces a novel structured random matrix composed blockwise from subsampled randomized Hadamard transforms (SRHTs). The block SRHT is expected to outperform well-known dimension reduction maps, including SRHT and Gaussian matrices on distributed architectures. We prove that a block SRHT with enough rows...	[]	null	6,779	2210.11295	title_judge
2K2vEVBm5G	Unconstrained Online Learning with Unbounded Losses	https://openreview.net/forum?id=2K2vEVBm5G	[ "Andrew Jacobsen", "Ashok Cutkosky" ]	Poster	null	Algorithms for online learning typically require one or more boundedness assumptions: that the domain is bounded, that the losses are Lipschitz, or both. In this paper, we develop a new setting for online learning with unbounded domains and non-Lipschitz losses. For this setting we provide an algorithm which guarantees...	[]	null	6,766	2306.04923	title_snapshot
LctoTBcGUf	Optimistic Planning by Regularized Dynamic Programming	https://openreview.net/forum?id=LctoTBcGUf	[ "Antoine Moulin", "Gergely Neu" ]	Poster	null	We propose a new method for optimistic planning in infinite-horizon discounted Markov decision processes based on the idea of adding regularization to the updates of an otherwise standard approximate value iteration procedure. This technique allows us to avoid contraction and monotonicity arguments typically required b...	[]	null	6,761	2302.14004	title_snapshot
dtg76BRYCG	Autoregressive Diffusion Model for Graph Generation	https://openreview.net/forum?id=dtg76BRYCG	[ "Lingkai Kong", "Jiaming Cui", "Haotian Sun", "Yuchen Zhuang", "B. Aditya Prakash", "Chao Zhang" ]	Poster	null	Diffusion-based graph generative models have recently obtained promising results for graph generation. However, existing diffusion-based graph generative models are mostly one-shot generative models that apply Gaussian diffusion in the dequantized adjacency matrix space. Such a strategy can suffer from difficulty in mo...	[]	null	6,751	2307.08849	title_snapshot
YZoYYaawO2	Differentiable Tree Operations Promote Compositional Generalization	https://openreview.net/forum?id=YZoYYaawO2	[ "Paul Soulos", "Edward J Hu", "Kate McCurdy", "Yunmo Chen", "Roland Fernandez", "Paul Smolensky", "Jianfeng Gao" ]	Poster	null	In the context of structure-to-structure transformation tasks, learning sequences of discrete symbolic operations poses significant challenges due to their non-differentiability. To facilitate the learning of these symbolic sequences, we introduce a differentiable tree interpreter that compiles high-level symbolic tree...	[]	null	6,739	2306.00751	title_snapshot
Pbaiy3fRCt	Can Neural Network Memorization Be Localized?	https://openreview.net/forum?id=Pbaiy3fRCt	[ "Pratyush Maini", "Michael Curtis Mozer", "Hanie Sedghi", "Zachary Chase Lipton", "J Zico Kolter", "Chiyuan Zhang" ]	Poster	null	Recent efforts at explaining the interplay of memorization and generalization in deep overparametrized networks have posited that neural networks memorize ``hard'' examples in the final few layers of the model. Memorization refers to the ability to correctly predict on atypical examples of the training set. In this...	[]	null	6,724	2307.09542	title_snapshot
tRhQsHnoFw	Bayesian Design Principles for Frequentist Sequential Learning	https://openreview.net/forum?id=tRhQsHnoFw	[ "Yunbei Xu", "assaf zeevi" ]	Oral	null	We develop a general theory to optimize the frequentist regret for sequential learning problems, where efficient bandit and reinforcement learning algorithms can be derived from unified Bayesian principles. We propose a novel optimization approach to create "algorithmic beliefs" at each round, and use Bayesian posterio...	[]	null	6,716	2310.00806	title_snapshot
Xzfur8Blaf	Domain Adaptation for Time Series Under Feature and Label Shifts	https://openreview.net/forum?id=Xzfur8Blaf	[ "Huan He", "Owen Queen", "Teddy Koker", "Consuelo Cuevas", "Theodoros Tsiligkaridis", "Marinka Zitnik" ]	Poster	null	Unsupervised domain adaptation (UDA) enables the transfer of models trained on source domains to unlabeled target domains. However, transferring complex time series models presents challenges due to the dynamic temporal structure variations across domains. This leads to feature shifts in the time and frequency represen...	[]	null	6,694	2302.03133	title_snapshot
ASOCqTnWIY	Towards Sustainable Learning: Coresets for Data-efficient Deep Learning	https://openreview.net/forum?id=ASOCqTnWIY	[ "Yu Yang", "Kang Hao", "Baharan Mirzasoleiman" ]	Poster	null	To improve the efficiency and sustainability of learning deep models, we propose CREST, the first scalable framework with rigorous theoretical guarantees to identify the most valuable examples for training non-convex models, particularly deep networks. To guarantee convergence to a stationary point of a non-convex func...	[]	null	6,690	2306.01244	title_snapshot
uIOw2ZE1U8	On Enhancing Expressive Power via Compositions of Single Fixed-Size ReLU Network	https://openreview.net/forum?id=uIOw2ZE1U8	[ "Shijun Zhang", "Jianfeng Lu", "Hongkai Zhao" ]	Poster	null	This paper explores the expressive power of deep neural networks through the framework of function compositions. We demonstrate that the repeated compositions of a single fixed-size ReLU network exhibit surprising expressive power, despite the limited expressive capabilities of the individual network itself. Specifical...	[]	null	6,689	2301.12353	title_snapshot
bF1LVbP493	Pix2Struct: Screenshot Parsing as Pretraining for Visual Language Understanding	https://openreview.net/forum?id=bF1LVbP493	[ "Kenton Lee", "Mandar Joshi", "Iulia Raluca Turc", "Hexiang Hu", "Fangyu Liu", "Julian Martin Eisenschlos", "Urvashi Khandelwal", "Peter Shaw", "Ming-Wei Chang", "Kristina Toutanova" ]	Oral	null	Visually-situated language is ubiquitous---sources range from textbooks with diagrams to web pages with images and tables, to mobile apps with buttons and forms. Perhaps due to this diversity, previous work has typically relied on domain-specific recipes with limited sharing of the underlying data, model architectures,...	[]	null	6,684	2210.03347	title_snapshot
ORxBEWMPAJ	JAWS-X: Addressing Efficiency Bottlenecks of Conformal Prediction Under Standard and Feedback Covariate Shift	https://openreview.net/forum?id=ORxBEWMPAJ	[ "Drew Prinster", "Suchi Saria", "Anqi Liu" ]	Oral	null	We study the efficient estimation of predictive confidence intervals for black-box predictors when the common data exchangeability (e.g., i.i.d.) assumption is violated due to potentially feedback-induced shifts in the input data distribution. That is, we focus on standard and feedback covariate shift (FCS), where the ...	[]	null	6,673	null	null
s1hrcLUcld	Contextual Reliability: When Different Features Matter in Different Contexts	https://openreview.net/forum?id=s1hrcLUcld	[ "Gaurav Rohit Ghosal", "Amrith Setlur", "Daniel S. Brown", "Anca Dragan", "Aditi Raghunathan" ]	Poster	null	Deep neural networks often fail catastrophically by relying on spurious correlations. Most prior work assumes a clear dichotomy into spurious and reliable features; however, this is often unrealistic. For example, most of the time we do not want an autonomous car to simply copy the speed of surrounding cars---we don't ...	[]	null	6,672	2307.10026	title_snapshot
XAGr6u76Lu	On Data Manifolds Entailed by Structural Causal Models	https://openreview.net/forum?id=XAGr6u76Lu	[ "Ricardo Dominguez-Olmedo", "Amir-Hossein Karimi", "Georgios Arvanitidis", "Bernhard Schölkopf" ]	Poster	null	The geometric structure of data is an important inductive bias in machine learning. In this work, we characterize the data manifolds entailed by structural causal models. The strengths of the proposed framework are twofold: firstly, the geometric structure of the data manifolds is causally informed, and secondly, it en...	[]	null	6,671	null	null
lwodnXJzu6	Are Neurons Actually Collapsed? On the Fine-Grained Structure in Neural Representations	https://openreview.net/forum?id=lwodnXJzu6	[ "Yongyi Yang", "Jacob Steinhardt", "Wei Hu" ]	Poster	null	Recent work has observed an intriguing "Neural Collapse'' phenomenon in well-trained neural networks, where the last-layer representations of training samples with the same label collapse into each other. This appears to suggest that the last-layer representations are completely determined by the labels, and do not dep...	[]	null	6,668	2306.17105	title_snapshot
mGUJMqjDwE	Provably Learning Object-Centric Representations	https://openreview.net/forum?id=mGUJMqjDwE	[ "Jack Brady", "Roland S. Zimmermann", "Yash Sharma", "Bernhard Schölkopf", "Julius von Kügelgen", "Wieland Brendel" ]	Oral	null	Learning structured representations of the visual world in terms of objects promises to significantly improve the generalization abilities of current machine learning models. While recent efforts to this end have shown promising empirical progress, a theoretical account of when unsupervised object-centric representatio...	[]	null	6,664	2305.14229	title_snapshot
gWC3Q3pyHe	Fast Sampling of Diffusion Models via Operator Learning	https://openreview.net/forum?id=gWC3Q3pyHe	[ "Hongkai Zheng", "Weili Nie", "Arash Vahdat", "Kamyar Azizzadenesheli", "Anima Anandkumar" ]	Poster	null	Diffusion models have found widespread adoption in various areas. However, their sampling process is slow because it requires hundreds to thousands of network evaluations to emulate a continuous process defined by differential equations. In this work, we use neural operators, an efficient method to solve the probabilit...	[]	null	6,663	2211.13449	title_snapshot
HiX1ybkFMl	Unsupervised Out-of-Distribution Detection with Diffusion Inpainting	https://openreview.net/forum?id=HiX1ybkFMl	[ "Zhenzhen Liu", "Jin Peng Zhou", "Yufan Wang", "Kilian Q Weinberger" ]	Poster	null	Unsupervised out-of-distribution detection (OOD) seeks to identify out-of-domain data by learning only from unlabeled in-domain data. We present a novel approach for this task -- Lift, Map, Detect (LMD) -- that leverages recent advancement in diffusion models. Diffusion models are one type of generative models. At thei...	[]	null	6,661	2302.10326	title_snapshot
PKrDN23Rke	Sequence Modeling with Multiresolution Convolutional Memory	https://openreview.net/forum?id=PKrDN23Rke	[ "Jiaxin Shi", "Ke Alexander Wang", "Emily Fox" ]	Poster	null	Efficiently capturing the long-range patterns in sequential data sources salient to a given task---such as classification and generative modeling---poses a fundamental challenge. Popular approaches in the space tradeoff between the memory burden of brute-force enumeration and comparison, as in transformers, the computa...	[]	null	6,657	2305.01638	title_snapshot
BrOPvKsIXW	The Hessian perspective into the Nature of Convolutional Neural Networks	https://openreview.net/forum?id=BrOPvKsIXW	[ "Sidak Pal Singh", "Thomas Hofmann", "Bernhard Schölkopf" ]	Poster	null	While Convolutional Neural Networks (CNNs) have long been investigated and applied, as well as theorized, we aim to provide a slightly different perspective into their nature --- through the perspective of their Hessian maps. The reason is that the loss Hessian captures the pairwise interaction of parameters and theref...	[]	null	6,656	2305.09088	title_snapshot
mSslPmao9h	Towards Learning Geometric Eigen-Lengths Crucial for Fitting Tasks	https://openreview.net/forum?id=mSslPmao9h	[ "Yijia Weng", "Kaichun Mo", "Ruoxi Shi", "Yanchao Yang", "Leonidas Guibas" ]	Poster	null	Some extremely low-dimensional yet crucial geometric eigen-lengths often determine the success of some geometric tasks. For example, the height of an object is important to measure to check if it can fit between the shelves of a cabinet, while the width of a couch is crucial when trying to move it through a doorway...	[]	null	6,651	2312.15610	title_snapshot
OxkESnZnN2	Improved Learning-Augmented Algorithms for the Multi-Option Ski Rental Problem via Best-Possible Competitive Analysis	https://openreview.net/forum?id=OxkESnZnN2	[ "Yongho Shin", "Changyeol Lee", "Gukryeol Lee", "Hyung-Chan An" ]	Poster	null	In this paper, we present improved learning-augmented algorithms for the multi-option ski rental problem. Learning-augmented algorithms take ML predictions as an added part of the input and incorporates these predictions in solving the given problem. Due to their unique strength that combines the power of ML prediction...	[]	null	6,642	2302.06832	title_snapshot
PsQJm6lG3s	On Regularization and Inference with Label Constraints	https://openreview.net/forum?id=PsQJm6lG3s	[ "Kaifu Wang", "Hangfeng He", "Tin D. Nguyen", "Piyush Kumar", "Dan Roth" ]	Poster	null	Prior knowledge and symbolic rules in machine learning are often expressed in the form of label constraints, especially in structured prediction problems. In this work, we compare two common strategies for encoding label constraints in a machine learning pipeline, regularization with constraints and *constrained infe...	[]	null	6,635	2307.03886	title_snapshot
oupdxuURWD	Simple Disentanglement of Style and Content in Visual Representations	https://openreview.net/forum?id=oupdxuURWD	[ "Lilian Ngweta", "Subha Maity", "Alex Gittens", "Yuekai Sun", "Mikhail Yurochkin" ]	Poster	null	Learning visual representations with interpretable features, i.e., disentangled representations, remains a challenging problem. Existing methods demonstrate some success but are hard to apply to large-scale vision datasets like ImageNet. In this work, we propose a simple post-processing framework to disentangle content...	[]	null	6,630	2302.09795	title_snapshot
AvwlrX9AQr	Beyond the Edge of Stability via Two-step Gradient Updates	https://openreview.net/forum?id=AvwlrX9AQr	[ "Lei Chen", "Joan Bruna" ]	Poster	null	Gradient Descent (GD) is a powerful workhorse of modern machine learning thanks to its scalability and efficiency in high-dimensional spaces. Its ability to find local minimisers is only guaranteed for losses with Lipschitz gradients, where it can be seen as a 'bona-fide' discretisation of an underlying gradient flow. ...	[]	null	6,628	2206.04172	title_snapshot
qorOnDor89	On the Role of Attention in Prompt-tuning	https://openreview.net/forum?id=qorOnDor89	[ "Samet Oymak", "Ankit Singh Rawat", "Mahdi Soltanolkotabi", "Christos Thrampoulidis" ]	Poster	null	Prompt-tuning is an emerging strategy to adapt large language models (LLM) to downstream tasks by learning a (soft-)prompt parameter from data. Despite its success in LLMs, there is limited theoretical understanding of the power of prompt-tuning and the role of the attention mechanism in prompting. In this work, we exp...	[]	null	6,614	2306.03435	title_snapshot
nrSM4XmF5k	Master-ASR: Achieving Multilingual Scalability and Low-Resource Adaptation in ASR with Modular Learning	https://openreview.net/forum?id=nrSM4XmF5k	[ "Zhongzhi Yu", "Yang Zhang", "Kaizhi Qian", "Cheng Wan", "Yonggan Fu", "Yongan Zhang", "Celine Lin" ]	Poster	null	Despite the impressive performance recently achieved by automatic speech recognition (ASR), we observe two primary challenges that hinder its broader applications: (1) The difficulty of introducing scalability into the model to support more languages with limited training, inference, and storage overhead; (2) The low-r...	[]	null	6,612	2306.15686	title_snapshot
9CZZ8tIhSv	Hyperbolic Representation Learning: Revisiting and Advancing	https://openreview.net/forum?id=9CZZ8tIhSv	[ "Menglin Yang", "min zhou", "Zhitao Ying", "Yankai Chen", "Irwin King" ]	Poster	null	The non-Euclidean geometry of hyperbolic spaces has recently garnered considerable attention in the realm of representation learning. Current endeavors in hyperbolic representation largely presuppose that the underlying hierarchies can be automatically inferred and preserved through the adaptive optimization process. T...	[]	null	6,611	2306.09118	title_snapshot
4IzEmHLono	Learning Belief Representations for Partially Observable Deep RL	https://openreview.net/forum?id=4IzEmHLono	[ "Andrew Wang", "Andrew C Li", "Toryn Q. Klassen", "Rodrigo Toro Icarte", "Sheila A. McIlraith" ]	Poster	null	Many important real-world Reinforcement Learning (RL) problems involve partial observability and require policies with memory. Unfortunately, standard deep RL algorithms for partially observable settings typically condition on the full history of interactions and are notoriously difficult to train. We propose a novel d...	[]	null	6,603	null	null
Asrg2we3dP	MABe22: A Multi-Species Multi-Task Benchmark for Learned Representations of Behavior	https://openreview.net/forum?id=Asrg2we3dP	[ "Jennifer J. Sun", "Markus Marks", "Andrew Wesley Ulmer", "Dipam Chakraborty", "Brian Geuther", "Edward Hayes", "Heng Jia", "Vivek Kumar", "Sebastian Oleszko", "Zachary Partridge", "Milan Peelman", "Alice Robie", "Catherine E Schretter", "Keith Sheppard", "Chao Sun", "Param Uttarwar", ...	Poster	null	We introduce MABe22, a large-scale, multi-agent video and trajectory benchmark to assess the quality of learned behavior representations. This dataset is collected from a variety of biology experiments, and includes triplets of interacting mice (4.7 million frames video+pose tracking data, 10 million frames pose only),...	[]	null	6,594	2207.10553	title_snapshot
ccwSdYv1GI	Scaling Up Dataset Distillation to ImageNet-1K with Constant Memory	https://openreview.net/forum?id=ccwSdYv1GI	[ "Justin Cui", "Ruochen Wang", "Si Si", "Cho-Jui Hsieh" ]	Poster	null	Dataset Distillation is a newly emerging area that aims to distill large datasets into much smaller and highly informative synthetic ones to accelerate training and reduce storage. Among various dataset distillation methods, trajectory-matching-based methods (MTT) have achieved SOTA performance in many tasks, e.g., on ...	[]	null	6,583	2211.10586	title_snapshot
mernbGTe24	MultiAdam: Parameter-wise Scale-invariant Optimizer for Multiscale Training of Physics-informed Neural Networks	https://openreview.net/forum?id=mernbGTe24	[ "Jiachen Yao", "Chang Su", "Zhongkai Hao", "Songming Liu", "Hang Su", "Jun Zhu" ]	Poster	null	Physics-informed Neural Networks (PINNs) have recently achieved remarkable progress in solving Partial Differential Equations (PDEs) in various fields by minimizing a weighted sum of PDE loss and boundary loss. However, there are several critical challenges in the training of PINNs, including the lack of theoretical fr...	[]	null	6,572	2306.02816	title_snapshot
gJboa2IOua	Polynomial Preconditioning for Gradient Methods	https://openreview.net/forum?id=gJboa2IOua	[ "Nikita Doikov", "Anton Rodomanov" ]	Poster	null	We study first-order methods with preconditioning for solving structured convex optimization problems. We propose a new family of preconditioners generated by the symmetric polynomials. They provide the first-order optimization methods with a provable improvement of the condition number, cutting the gaps between highes...	[]	null	6,562	2301.13194	title_snapshot
lgEYRIk7GS	Internally Rewarded Reinforcement Learning	https://openreview.net/forum?id=lgEYRIk7GS	[ "Mengdi Li", "Xufeng Zhao", "Jae Hee Lee", "Cornelius Weber", "Stefan Wermter" ]	Poster	null	We study a class of reinforcement learning problems where the reward signals for policy learning are generated by a discriminator that is dependent on and jointly optimized with the policy. This interdependence between the policy and the discriminator leads to an unstable learning process because reward signals from an...	[]	null	6,561	2302.00270	title_snapshot
Lq5H6B6yug	Lottery Tickets in Evolutionary Optimization: On Sparse Backpropagation-Free Trainability	https://openreview.net/forum?id=Lq5H6B6yug	[ "Robert Tjarko Lange", "Henning Sprekeler" ]	Poster	null	Is the lottery ticket phenomenon an idiosyncrasy of gradient-based training or does it generalize to evolutionary optimization? In this paper we establish the existence of highly sparse trainable initializations for evolution strategies (ES) and characterize qualitative differences compared to gradient descent (GD)-bas...	[]	null	6,560	2306.00045	title_snapshot
ABzDOXlxf0	Slot-VAE: Object-Centric Scene Generation with Slot Attention	https://openreview.net/forum?id=ABzDOXlxf0	[ "Yanbo Wang", "Letao Liu", "Justin Dauwels" ]	Poster	null	Slot attention has shown remarkable object-centric representation learning performance in computer vision tasks without requiring any supervision. Despite its object-centric binding ability brought by compositional modelling, as a deterministic module, slot attention lacks the ability to generate novel scenes. In this ...	[]	null	6,552	2306.06997	title_snapshot
rDMAJECBM2	Model-based Reinforcement Learning with Scalable Composite Policy Gradient Estimators	https://openreview.net/forum?id=rDMAJECBM2	[ "Paavo Parmas", "Takuma Seno", "Yuma Aoki" ]	Poster	null	In model-based reinforcement learning (MBRL), policy gradients can be estimated either by derivative-free RL methods, such as likelihood ratio gradients (LR), or by backpropagating through a differentiable model via reparameterization gradients (RP). Instead of using one or the other, the Total Propagation (TP) algorit...	[]	null	6,550	null	null
MtopPVk3Ll	H-Likelihood Approach to Deep Neural Networks with Temporal-Spatial Random Effects for High-Cardinality Categorical Features	https://openreview.net/forum?id=MtopPVk3Ll	[ "Hangbin Lee", "Youngjo Lee" ]	Oral	null	Deep Neural Networks (DNNs) are one of the most powerful tools for prediction, but many of them implicitly assume that the data are statistically independent. However, in the real world, it is common for large-scale data to be clustered with temporal-spatial correlation structures. Variational approaches and integrated...	[]	null	6,548	null	null
uSJP34JCTu	Predicting Rare Events by Shrinking Towards Proportional Odds	https://openreview.net/forum?id=uSJP34JCTu	[ "Gregory Faletto", "Jacob Bien" ]	Poster	null	Training classifiers is difficult with severe class imbalance, but many rare events are the culmination of a sequence with much more common intermediate outcomes. For example, in online marketing a user first sees an ad, then may click on it, and finally may make a purchase; estimating the probability of purchases is d...	[]	null	6,542	2305.18700	title_snapshot
OT6gRRMmcE	The Optimal Approximation Factors in Misspecified Off-Policy Value Function Estimation	https://openreview.net/forum?id=OT6gRRMmcE	[ "Philip Amortila", "Nan Jiang", "Csaba Szepesvari" ]	Poster	null	Theoretical guarantees in reinforcement learning (RL) are known to suffer multiplicative blow-up factors with respect to the misspecification error of function approximation. Yet, the nature of such approximation factors---especially their optimal form in a given learning problem---is poorly understood. In this paper...	[]	null	6,539	2307.13332	title_snapshot
QBUS4OXqvM	Internet Explorer: Targeted Representation Learning on the Open Web	https://openreview.net/forum?id=QBUS4OXqvM	[ "Alexander Cong Li", "Ellis Langham Brown", "Alexei A Efros", "Deepak Pathak" ]	Poster	null	Vision models typically rely on fine-tuning general-purpose models pre-trained on large, static datasets. These general-purpose models only capture the knowledge within their pre-training datasets, which are tiny, out-of-date snapshots of the Internet---where billions of images are uploaded each day. We suggest an alte...	[]	null	6,537	2302.14051	title_snapshot
eYlLlvzngu	Using Large Language Models to Simulate Multiple Humans and Replicate Human Subject Studies	https://openreview.net/forum?id=eYlLlvzngu	[ "Gati Aher", "Rosa I. Arriaga", "Adam Tauman Kalai" ]	Oral	null	We introduce a new type of test, called a Turing Experiment (TE), for evaluating to what extent a given language model, such as GPT models, can simulate different aspects of human behavior. A TE can also reveal consistent distortions in a language model’s simulation of a specific human behavior. Unlike the Turing Test,...	[]	null	6,534	2208.10264	title_snapshot
r3M5cBtpYq	Robust and private stochastic linear bandits	https://openreview.net/forum?id=r3M5cBtpYq	[ "Vasileios Charisopoulos", "Hossein Esfandiari", "Vahab Mirrokni" ]	Poster	null	In this paper, we study the stochastic linear bandit problem under the additional requirements of differential privacy, robustness and batched observations. In particular, we assume an adversary randomly chooses a constant fraction of the observed rewards in each batch, replacing them with arbitrary numbers. We p...	[]	null	6,533	null	null
wGgIcftFzm	High-dimensional Location Estimation via Norm Concentration for Subgamma Vectors	https://openreview.net/forum?id=wGgIcftFzm	[ "Shivam Gupta", "Jasper C.H. Lee", "Eric Price" ]	Poster	null	In location estimation, we are given $n$ samples from a known distribution $f$ shifted by an unknown translation $\lambda$, and want to estimate $\lambda$ as precisely as possible. Asymptotically, the maximum likelihood estimate achieves the Cramér-Rao bound of error $\mathcal N(0, \frac{1}{n\mathcal I})$, where $\math...	[]	null	6,529	2302.02497	title_snapshot
E3BW8pG64Y	Action Matching: Learning Stochastic Dynamics from Samples	https://openreview.net/forum?id=E3BW8pG64Y	[ "Kirill Neklyudov", "Rob Brekelmans", "Daniel Severo", "Alireza Makhzani" ]	Poster	null	Learning the continuous dynamics of a system from snapshots of its temporal marginals is a problem which appears throughout natural sciences and machine learning, including in quantum systems, single-cell biological data, and generative modeling. In these settings, we assume access to cross-sectional samples that are u...	[]	null	6,526	2210.06662	title_snapshot
W5xbQluQ5d	Short-lived High-volume Bandits	https://openreview.net/forum?id=W5xbQluQ5d	[ "Su Jia", "Nishant Oli", "Ian Anderson", "Paul Duff", "Andrew A Li", "Ramamoorthi Ravi" ]	Poster	null	Modern platforms leverage randomized experiments to make informed decisions from a given set of alternatives. As a particularly challenging scenario, these alternatives can potentially have (i) high volume, with thousands of new items being released each hour, and (ii) short lifetime, either due to the contents' transi...	[]	null	6,522	null	null
YeTYJz7th5	Learning Temporally AbstractWorld Models without Online Experimentation	https://openreview.net/forum?id=YeTYJz7th5	[ "Benjamin Freed", "Siddarth Venkatraman", "Guillaume Adrien Sartoretti", "Jeff Schneider", "Howie Choset" ]	Poster	null	Agents that can build temporally abstract representations of their environment are better able to understand their world and make plans on extended time scales, with limited computational power and modeling capacity. However, existing methods for automatically learning temporally abstract world models usually require m...	[]	null	6,518	null	null
HAtMGUv1ci	Active Policy Improvement from Multiple Black-box Oracles	https://openreview.net/forum?id=HAtMGUv1ci	[ "Xuefeng Liu", "Takuma Yoneda", "Chaoqi Wang", "Matthew Walter", "Yuxin Chen" ]	Poster	null	Reinforcement learning (RL) has made significant strides in various complex domains. However, identifying an effective policy via RL often necessitates extensive exploration. Imitation learning aims to mitigate this issue by using expert demonstrations to guide exploration. In real-world scenarios, one often has access...	[]	null	6,517	2306.10259	title_snapshot
yqUhEFPoDN	Stochastic Gradient Descent-Induced Drift of Representation in a Two-Layer Neural Network	https://openreview.net/forum?id=yqUhEFPoDN	[ "Farhad Pashakhanloo", "Alexei Koulakov" ]	Poster	null	Representational drift refers to over-time changes in neural activation accompanied by a stable task performance. Despite being observed in the brain and in artificial networks, the mechanisms of drift and its implications are not fully understood. Motivated by recent experimental findings of stimulus-dependent drift i...	[]	null	6,498	2302.02563	title_snapshot
MZkbgahv4a	IRNeXt: Rethinking Convolutional Network Design for Image Restoration	https://openreview.net/forum?id=MZkbgahv4a	[ "Yuning Cui", "Wenqi Ren", "Sining Yang", "Xiaochun Cao", "Alois Knoll" ]	Poster	null	We present IRNeXt, a simple yet effective convolutional network architecture for image restoration. Recently, Transformer models have dominated the field of image restoration due to the powerful ability of modeling long-range pixels interactions. In this paper, we excavate the potential of the convolutional neural netw...	[]	null	6,492	null	null
mRiDy4qGwB	TabLeak: Tabular Data Leakage in Federated Learning	https://openreview.net/forum?id=mRiDy4qGwB	[ "Mark Vero", "Mislav Balunovic", "Dimitar Iliev Dimitrov", "Martin Vechev" ]	Poster	null	While federated learning (FL) promises to preserve privacy, recent works in the image and text domains have shown that training updates leak private client data. However, most high-stakes applications of FL (e.g., in healthcare and finance) use tabular data, where the risk of data leakage has not yet been explored. A s...	[]	null	6,491	2210.01785	title_snapshot
mXv2aVqUGG	Can Large Language Models Reason about Program Invariants?	https://openreview.net/forum?id=mXv2aVqUGG	[ "Kexin Pei", "David Bieber", "Kensen Shi", "Charles Sutton", "Pengcheng Yin" ]	Poster	null	Identifying invariants is an important program analysis task with applications towards program understanding, bug finding, vulnerability analysis, and formal verification. Existing tools for identifying program invariants rely on dynamic analysis, requiring traces collected from multiple executions in order to produce ...	[]	null	6,486	null	null
occOHdHWRn	Probabilistic Unrolling: Scalable, Inverse-Free Maximum Likelihood Estimation for Latent Gaussian Models	https://openreview.net/forum?id=occOHdHWRn	[ "Alexander Lin", "Bahareh Tolooshams", "Yves Atchade", "Demba E. Ba" ]	Poster	null	Latent Gaussian models have a rich history in statistics and machine learning, with applications ranging from factor analysis to compressed sensing to time series analysis. The classical method for maximizing the likelihood of these models is the expectation-maximization (EM) algorithm. For problems with high-dimension...	[]	null	6,480	2306.03249	title_snapshot
2DiRkdZJYm	Stable and Consistent Prediction of 3D Characteristic Orientation via Invariant Residual Learning	https://openreview.net/forum?id=2DiRkdZJYm	[ "Seungwook Kim", "Chunghyun Park", "Yoonwoo Jeong", "Jaesik Park", "Minsu Cho" ]	Poster	null	Learning to predict reliable characteristic orientations of 3D point clouds is an important yet challenging problem, as different point clouds of the same class may have largely varying appearances. In this work, we introduce a novel method to decouple the shape geometry and semantics of the input point cloud to achiev...	[]	null	6,479	2306.11406	title_snapshot
vDMHusV7J0	Deep Generative Symbolic Regression with Monte-Carlo-Tree-Search	https://openreview.net/forum?id=vDMHusV7J0	[ "Pierre-Alexandre Kamienny", "Guillaume Lample", "sylvain lamprier", "Marco Virgolin" ]	Poster	null	Symbolic regression (SR) is the problem of learning a symbolic expression from numerical data. Recently, deep neural models trained on procedurally-generated synthetic datasets showed competitive performance compared to more classical Genetic Programming (GP) ones. Unlike their GP counterparts, these neural approaches ...	[]	null	6,477	2302.11223	title_snapshot
HqQIt6mt5B	Mixing Predictions for Online Metric Algorithms	https://openreview.net/forum?id=HqQIt6mt5B	[ "Antonios Antoniadis", "Christian Coester", "Marek Elias", "Adam Polak", "Bertrand Simon" ]	Poster	null	A major technique in learning-augmented online algorithms is combining multiple algorithms or predictors. Since the performance of each predictor may vary over time, it is desirable to use not the single best predictor as a benchmark, but rather a dynamic combination which follows different predictors at different time...	[]	null	6,468	2304.01781	title_snapshot
iRBKUnIjR2	Approximate Causal Effect Identification under Weak Confounding	https://openreview.net/forum?id=iRBKUnIjR2	[ "Ziwei Jiang", "Lai Wei", "Murat Kocaoglu" ]	Poster	null	Causal effect estimation has been studied by many researchers when only observational data is available. Sound and complete algorithms have been developed for pointwise estimation of identifiable causal queries. For non-identifiable causal queries, researchers developed polynomial programs to estimate tight bounds on c...	[]	null	6,466	2306.13242	title_snapshot
RIf7TPG5H0	Bootstrap in High Dimension with Low Computation	https://openreview.net/forum?id=RIf7TPG5H0	[ "Henry Lam", "Zhenyuan Liu" ]	Poster	null	The bootstrap is a popular data-driven method to quantify statistical uncertainty, but for modern high-dimensional problems, it could suffer from huge computational costs due to the need to repeatedly generate resamples and refit models. We study the use of bootstraps in high-dimensional environments with a small numbe...	[]	null	6,465	2210.10974	title_snapshot
O4hhnt07Yk	Image Shortcut Squeezing: Countering Perturbative Availability Poisons with Compression	https://openreview.net/forum?id=O4hhnt07Yk	[ "Zhuoran Liu", "Zhengyu Zhao", "Martha Larson" ]	Poster	null	Perturbative availability poisoning (PAP) adds small changes to images to prevent their use for model training. Current research adopts the belief that practical and effective approaches to countering such poisons do not exist. In this paper, we argue that it is time to abandon this belief. We present extensive experim...	[]	null	6,459	2301.13838	title_snapshot
VVdb1la0cW	Proper Losses for Discrete Generative Models	https://openreview.net/forum?id=VVdb1la0cW	[ "Dhamma Kimpara", "Rafael Frongillo", "Bo Waggoner" ]	Poster	null	We initiate the study of proper losses for evaluating generative models in the discrete setting. Unlike traditional proper losses, we treat both the generative model and the target distribution as black-boxes, only assuming ability to draw i.i.d. samples. We define a loss to be black-box proper if the generative distri...	[]	null	6,454	2211.03761	title_snapshot
y8qAZhWbNs	Private Federated Learning with Autotuned Compression	https://openreview.net/forum?id=y8qAZhWbNs	[ "Enayat Ullah", "Christopher A. Choquette-Choo", "Peter Kairouz", "Sewoong Oh" ]	Poster	null	We propose new techniques for reducing communication in private federated learning without the need for setting or tuning compression rates. Our on-the-fly methods automatically adjust the compression rate based on the error induced during training, while maintaining provable privacy guarantees through the use of secur...	[]	null	6,453	2307.10999	title_snapshot
vOcOzRWpvm	SpotEM: Efficient Video Search for Episodic Memory	https://openreview.net/forum?id=vOcOzRWpvm	[ "Santhosh Kumar Ramakrishnan", "Ziad Al-Halah", "Kristen Grauman" ]	Poster	null	The goal in episodic memory (EM) is to search a long egocentric video to answer a natural language query (e.g., “where did I leave my purse?”). Existing EM methods exhaustively extract expensive fixed-length clip features to look everywhere in the video for the answer, which is infeasible for long wearable-camera video...	[]	null	6,452	2306.15850	title_snapshot
JrSWhb7dzp	DRCFS: Doubly Robust Causal Feature Selection	https://openreview.net/forum?id=JrSWhb7dzp	[ "Francesco Quinzan", "Ashkan Soleymani", "Patrick Jaillet", "Cristian R. Rojas", "Stefan Bauer" ]	Poster	null	Knowing the features of a complex system that are highly relevant to a particular target variable is of fundamental interest in many areas of science. Existing approaches are often limited to linear settings, sometimes lack guarantees, and in most cases, do not scale to the problem at hand, in particular to images. We ...	[]	null	6,451	2306.07024	title_snapshot
ET6qkbzeOx	Tractable Control for Autoregressive Language Generation	https://openreview.net/forum?id=ET6qkbzeOx	[ "Honghua Zhang", "Meihua Dang", "Nanyun Peng", "Guy Van den Broeck" ]	Oral	null	Despite the success of autoregressive large language models in text generation, it remains a major challenge to generate text that satisfies complex constraints: sampling from the conditional distribution ${\Pr}(\text{text} \| \alpha)$ is intractable for even the simplest lexical constraints $\alpha$. To overcome this c...	[]	null	6,436	2304.07438	title_snapshot
Hsfchgv3WW	Open-Vocabulary Universal Image Segmentation with MaskCLIP	https://openreview.net/forum?id=Hsfchgv3WW	[ "Zheng Ding", "Jieke Wang", "Zhuowen Tu" ]	Poster	null	In this paper, we tackle an emerging computer vision task, open-vocabulary universal image segmentation, that aims to perform semantic/instance/panoptic segmentation (background semantic labeling + foreground instance segmentation) for arbitrary categories of text-based descriptions in inference time. We first build a ...	[]	null	6,434	2208.08984	title_snapshot
FREvWGzoRu	Universal Physics-Informed Neural Networks: Symbolic Differential Operator Discovery with Sparse Data	https://openreview.net/forum?id=FREvWGzoRu	[ "Lena Podina", "Brydon Eastman", "Mohammad Kohandel" ]	Poster	null	In this work we perform symbolic discovery of differential operators in a situation where there is sparse experimental data. This small data regime in machine learning can be made tractable by providing our algorithms with prior information about the underlying dynamics. Physics Informed Neural Networks (PINNs) have be...	[]	null	6,431	2212.04630	title_judge
kdrPtUAfNx	Partial Optimality in Cubic Correlation Clustering	https://openreview.net/forum?id=kdrPtUAfNx	[ "David Stein", "Silvia Di Gregorio", "Bjoern Andres" ]	Poster	null	The higher-order correlation clustering problem is an expressive model, and recently, local search heuristics have been proposed for several applications. Certifying optimality, however, is NP-hard and practically hampered already by the complexity of the problem statement. Here, we focus on establishing partial optima...	[]	null	6,430	2302.04694	title_snapshot
ycZSQdo2F9	InfoDiffusion: Representation Learning Using Information Maximizing Diffusion Models	https://openreview.net/forum?id=ycZSQdo2F9	[ "Yingheng Wang", "Yair Schiff", "Aaron Gokaslan", "Weishen Pan", "Fei Wang", "Christopher De Sa", "Volodymyr Kuleshov" ]	Poster	null	While diffusion models excel at generating high-quality samples, their latent variables typically lack semantic meaning and are not suitable for representation learning. Here, we propose InfoDiffusion, an algorithm that augments diffusion models with low-dimensional latent variables that capture high-level factors of v...	[]	null	6,425	2306.08757	title_snapshot
3fNVNNyKyV	A Large-Scale Study of Probabilistic Calibration in Neural Network Regression	https://openreview.net/forum?id=3fNVNNyKyV	[ "Victor Dheur", "Souhaib Ben Taieb" ]	Poster	null	Accurate probabilistic predictions are essential for optimal decision making. While neural network miscalibration has been studied primarily in classification, we investigate this in the less-explored domain of regression. We conduct the largest empirical study to date to assess the probabilistic calibration of neural ...	[]	null	6,411	2306.02738	title_snapshot
szQzz2H8er	Global optimality of Elman-type RNNs in the mean-field regime	https://openreview.net/forum?id=szQzz2H8er	[ "Andrea Agazzi", "Jianfeng Lu", "Sayan Mukherjee" ]	Poster	null	We analyze Elman-type recurrent neural networks (RNNs) and their training in the mean-field regime. Specifically, we show convergence of gradient descent training dynamics of the RNN to the corresponding mean-field formulation in the large width limit. We also show that the fixed points of the limiting infinite-width d...	[]	null	6,410	2303.06726	title_judge
Skrk3StS2g	Semi-Autoregressive Energy Flows: Exploring Likelihood-Free Training of Normalizing Flows	https://openreview.net/forum?id=Skrk3StS2g	[ "Phillip Si", "Zeyi Chen", "Subham Sekhar Sahoo", "Yair Schiff", "Volodymyr Kuleshov" ]	Poster	null	Training normalizing flow generative models can be challenging due to the need to calculate computationally expensive determinants of Jacobians. This paper studies the likelihood-free training of flows and proposes the energy objective, an alternative sample-based loss based on proper scoring rules. The energy objectiv...	[]	null	6,409	2206.06672	title_snapshot
bZXfHpbUFi	When do Minimax-fair Learning and Empirical Risk Minimization Coincide?	https://openreview.net/forum?id=bZXfHpbUFi	[ "Harvineet Singh", "Matthäus Kleindessner", "Volkan Cevher", "Rumi Chunara", "Chris Russell" ]	Poster	null	Minimax-fair machine learning minimizes the error for the worst-off group. However, empirical evidence suggests that when sophisticated models are trained with standard empirical risk minimization (ERM), they often have the same performance on the worst-off group as a minimax-trained model. Our work makes this counter-...	[]	null	6,404	null	null
BdwGV6fwbK	Theoretical Behavior of XAI Methods in the Presence of Suppressor Variables	https://openreview.net/forum?id=BdwGV6fwbK	[ "Rick Wilming", "Leo Kieslich", "Benedict Clark", "Stefan Haufe" ]	Poster	null	In recent years, the community of 'explainable artificial intelligence' (XAI) has created a vast body of methods to bridge a perceived gap between model 'complexity' and 'interpretability'. However, a concrete problem to be solved by XAI methods has not yet been formally stated. As a result, XAI methods are lacking the...	[]	null	6,401	2306.01464	title_snapshot
7pcZLgulIV	Probabilistic Imputation for Time-series Classification with Missing Data	https://openreview.net/forum?id=7pcZLgulIV	[ "SeungHyun Kim", "Hyunsu Kim", "Eunggu Yun", "Hwangrae Lee", "Jaehun Lee", "Juho Lee" ]	Poster	null	Multivariate time series data for real-world applications typically contain a significant amount of missing values. The dominant approach for classification with such missing values is to impute them heuristically with specific values (zero, mean, values of adjacent time-steps) or learnable parameters. However, these s...	[]	null	6,400	2308.06738	title_snapshot
khats8j30c	MAHALO: Unifying Offline Reinforcement Learning and Imitation Learning from Observations	https://openreview.net/forum?id=khats8j30c	[ "Anqi Li", "Byron Boots", "Ching-An Cheng" ]	Poster	null	We study a new paradigm for sequential decision making, called offline policy learning from observations (PLfO). Offline PLfO aims to learn policies using datasets with substandard qualities: 1) only a subset of trajectories is labeled with rewards, 2) labeled trajectories may not contain actions, 3) labeled trajectori...	[]	null	6,399	2303.17156	title_snapshot
gzjK23oK9i	Robust Counterfactual Explanations for Neural Networks With Probabilistic Guarantees	https://openreview.net/forum?id=gzjK23oK9i	[ "Faisal Hamman", "Erfaun Noorani", "Saumitra Mishra", "Daniele Magazzeni", "Sanghamitra Dutta" ]	Poster	null	There is an emerging interest in generating robust counterfactual explanations that would remain valid if the model is updated or changed even slightly. Towards finding robust counterfactuals, existing literature often assumes that the original model $m$ and the new model $M$ are bounded in the parameter space, i.e., $...	[]	null	6,389	2305.11997	title_snapshot
IJffiJTLhI	Oracles & Followers: Stackelberg Equilibria in Deep Multi-Agent Reinforcement Learning	https://openreview.net/forum?id=IJffiJTLhI	[ "Matthias Gerstgrasser", "David C. Parkes" ]	Poster	null	Stackelberg equilibria arise naturally in a range of popular learning problems, such as in security games or indirect mechanism design, and have received increasing attention in the reinforcement learning literature. We present a general framework for implementing Stackelberg equilibria search as a multi-agent RL probl...	[]	null	6,387	2210.11942	title_snapshot
LVluQl5lAk	Multi-Agent Learning from Learners	https://openreview.net/forum?id=LVluQl5lAk	[ "Mine Melodi Caliskan", "Francesco Chini", "Setareh Maghsudi" ]	Poster	null	A large body of the "Inverse Reinforcement Learning" (IRL) literature focuses on recovering the reward function from a set of demonstrations of an expert agent who acts optimally or noisily optimally. Nevertheless, some recent works move away from the optimality assumption to study the "Learning from a Learner (LfL)" p...	[]	null	6,385	null	null
teYGEHBSYC	Cut your Losses with Squentropy	https://openreview.net/forum?id=teYGEHBSYC	[ "Like Hui", "Mikhail Belkin", "Stephen Wright" ]	Poster	null	Nearly all practical neural models for classification are trained using the cross-entropy loss. Yet this ubiquitous choice is supported by little theoretical or empirical evidence. Recent work (Hui & Belkin, 2020) suggests that training using the (rescaled) square loss is often superior in terms of the classification a...	[]	null	6,382	2302.03952	title_snapshot
tyqL1bPl0L	The Statistical Scope of Multicalibration	https://openreview.net/forum?id=tyqL1bPl0L	[ "Georgy Noarov", "Aaron Roth" ]	Poster	null	We make a connection between multicalibration and property elicitation and show that (under mild technical conditions) it is possible to produce a multicalibrated predictor for a continuous scalar property $\Gamma$ if and only if $\Gamma$ is elicitable. On the negative side, we show that for non-elicitable continuous...	[]	null	6,380	null	null
i7ZqmxsJTW	Sparse Learning of Dynamical Systems in RKHS: An Operator-Theoretic Approach	https://openreview.net/forum?id=i7ZqmxsJTW	[ "Boya Hou", "Sina Sanjari", "Nathan Dahlin", "Subhonmesh Bose", "Umesh Vaidya" ]	Poster	null	Transfer operators provide a rich framework for representing the dynamics of very general, nonlinear dynamical systems. When interacting with reproducing kernel Hilbert spaces (RKHS), descriptions of dynamics often incur prohibitive data storage requirements, motivating dataset sparsification as a precursory step to co...	[]	null	6,374	null	null
BUv0BLrosh	Conformal Prediction with Missing Values	https://openreview.net/forum?id=BUv0BLrosh	[ "Margaux Zaffran", "Aymeric Dieuleveut", "Julie Josse", "Yaniv Romano" ]	Poster	null	Conformal prediction is a theoretically grounded framework for constructing predictive intervals. We study conformal prediction with missing values in the covariates -- a setting that brings new challenges to uncertainty quantification. We first show that the marginal coverage guarantee of conformal prediction holds on...	[]	null	6,373	2306.02732	title_snapshot
O2XerBwfFk	Weakly Supervised Regression with Interval Targets	https://openreview.net/forum?id=O2XerBwfFk	[ "Xin Cheng", "Yuzhou Cao", "Ximing Li", "Bo An", "Lei Feng" ]	Poster	null	This paper investigates an interesting weakly supervised regression setting called regression with interval targets (RIT). Although some of the previous methods on relevant regression settings can be adapted to RIT, they are not statistically consistent, and thus their empirical performance is not guaranteed. In this p...	[]	null	6,358	2306.10458	title_snapshot
EiHX7MfAG0	Controllable Neural Symbolic Regression	https://openreview.net/forum?id=EiHX7MfAG0	[ "Tommaso Bendinelli", "Luca Biggio", "Pierre-Alexandre Kamienny" ]	Poster	null	In symbolic regression, the objective is to find an analytical expression that accurately fits experimental data with the minimal use of mathematical symbols such as operators, variables, and constants. However, the combinatorial space of possible expressions can make it challenging for traditional evolutionary algorit...	[]	null	6,353	2304.10336	title_snapshot
nVO6YTca8O	DSGD-CECA: Decentralized SGD with Communication-Optimal Exact Consensus Algorithm	https://openreview.net/forum?id=nVO6YTca8O	[ "Lisang Ding", "Kexin Jin", "Bicheng Ying", "Kun Yuan", "Wotao Yin" ]	Poster	null	Decentralized Stochastic Gradient Descent (SGD) is an emerging neural network training approach that enables multiple agents to train a model collaboratively and simultaneously. Rather than using a central parameter server to collect gradients from all the agents, each agent keeps a copy of the model parameters and com...	[]	null	6,346	2306.00256	title_snapshot
LMXgU4zrq6	How Do Transformers Learn Topic Structure: Towards a Mechanistic Understanding	https://openreview.net/forum?id=LMXgU4zrq6	[ "Yuchen Li", "Yuanzhi Li", "Andrej Risteski" ]	Poster	null	While the successes of transformers across many domains are indisputable, accurate understanding of the learning mechanics is still largely lacking. Their capabilities have been probed on benchmarks which include a variety of structured and reasoning tasks---but mathematical understanding is lagging substantially behin...	[]	null	6,342	2303.04245	title_snapshot
450iImFM4U	Self-Repellent Random Walks on General Graphs - Achieving Minimal Sampling Variance via Nonlinear Markov Chains	https://openreview.net/forum?id=450iImFM4U	[ "Vishwaraj Doshi", "Jie Hu", "Do Young Eun" ]	Oral	null	We consider random walks on discrete state spaces, such as general undirected graphs, where the random walkers are designed to approximate a target quantity over the network topology via sampling and neighborhood exploration in the form of Markov chain Monte Carlo (MCMC) procedures. Given any Markov chain corresponding...	[]	null	6,337	2305.05097	title_snapshot
nm4NwFfp7a	Subset-Based Instance Optimality in Private Estimation	https://openreview.net/forum?id=nm4NwFfp7a	[ "Travis Dick", "Alex Kulesza", "Ziteng Sun", "Ananda Theertha Suresh" ]	Poster	null	We propose a new definition of instance optimality for differentially private estimation algorithms. Our definition requires an optimal algorithm to compete, simultaneously for every dataset $D$, with the best private benchmark algorithm that (a) knows $D$ in advance and (b) is evaluated by its worst-case performance o...	[]	null	6,336	2303.01262	title_snapshot
0bR5JuxaoN	A Statistical Perspective on Retrieval-Based Models	https://openreview.net/forum?id=0bR5JuxaoN	[ "Soumya Basu", "Ankit Singh Rawat", "Manzil Zaheer" ]	Poster	null	Many modern high-performing machine learning models increasingly rely on scaling up models, e.g., transformer networks. Simultaneously, a parallel line of work aims to improve the model performance by augmenting an input instance with other (labeled) instances during inference. Examples of such augmentations include ta...	[]	null	6,333	null	null
CNq0JvrDfw	Adaptive IMLE for Few-shot Pretraining-free Generative Modelling	https://openreview.net/forum?id=CNq0JvrDfw	[ "Mehran Aghabozorgi", "Shichong Peng", "Ke Li" ]	Poster	null	Despite their success on large datasets, GANs have been difficult to apply in the few-shot setting, where only a limited number of training examples are provided. Due to mode collapse, GANs tend to ignore some training examples, causing overfitting to a subset of the training dataset, which is small in the first place....	[]	null	6,328	null	null
F2OjOG4j55	Polyhedral Complex Extraction from ReLU Networks using Edge Subdivision	https://openreview.net/forum?id=F2OjOG4j55	[ "Arturs Berzins" ]	Poster	null	A neural network consisting of piecewise affine building blocks, such as fully-connected layers and ReLU activations, is itself a piecewise affine function supported on a polyhedral complex. This complex has been previously studied to characterize theoretical properties of neural networks, but, in practice, extracting ...	[]	null	6,326	2306.07212	title_snapshot
Rgnaj43Pk0	Task-Specific Skill Localization in Fine-tuned Language Models	https://openreview.net/forum?id=Rgnaj43Pk0	[ "Abhishek Panigrahi", "Nikunj Saunshi", "Haoyu Zhao", "Sanjeev Arora" ]	Poster	null	Pre-trained language models can be fine-tuned to solve diverse NLP tasks, including in few-shot settings. Thus fine-tuning allows the model to quickly pick up task-specific "skills," but there has been limited study of where these newly-learnt skills reside inside the massive model. This paper introduces the term *sk...	[]	null	6,325	2302.06600	title_snapshot

End of preview. Expand in Data Studio

README.md exists but content is empty.

Downloads last month: 46

Collection including ai-conferences/ICML2023

ICML

Collection

Accepted papers for ICML (International Conference on Machine Learning), one dataset per year. • 14 items • Updated 3 days ago