Search Results for author:
Found 66 papers, 19 papers with code
EmbedLLM: Learning Compact Representations of Large Language Models
Many existing methods repeatedly learn task-specific representations of Large Language Models (LLMs), which leads to inefficiencies in both time and computational resources.
Looped Transformers for Length Generalization
Recent work has shown that Transformers trained from scratch can successfully solve various arithmetic and algorithmic tasks, such as adding numbers and computing parity.
Transformers on Markov Data: Constant Depth Suffices
On the theoretical side, our main result is that a transformer with a single head and three layers can represent the in-context conditional empirical distribution for \kth Markov sources, concurring with our empirical observations.
Toward a Theory of Tokenization in LLMs
In this paper, we investigate tokenization from a theoretical point of view by studying the behavior of transformers on simple data generating processes.
Learning to Understand: Identifying Interactions via the Möbius Transform
This marks the first $n$ sub-linear query complexity, noise-tolerant algorithm for the M\"obius transform.
Towards Optimal Statistical Watermarking
Key to our formulation is a coupling of the output tokens and the rejection region, realized by pseudo-random generators in practice, that allows non-trivial trade-offs between the Type I error and Type II error.
Pairwise Proximal Policy Optimization: Harnessing Relative Feedback for LLM Alignment
In summary, this work introduces a simpler yet effective approach for aligning LLMs to human preferences through relative feedback.
MRI Reconstruction with Side Information using Diffusion Models
This leads to an improvement in image reconstruction fidelity over generative models that rely only on a marginal prior over the image contrast of interest.
Statistical Complexity and Optimal Algorithms for Non-linear Ridge Bandits
We consider the sequential decision-making problem where the mean outcome is a non-linear function of the chosen action.
The Fair Value of Data Under Heterogeneous Privacy Constraints in Federated Learning
We also formulate a heterogeneous federated learning problem for the platform with privacy level options for users.
Efficiently Computing Sparse Fourier Transforms of $q$-ary Functions
Fourier transformations of pseudo-Boolean functions are popular tools for analyzing functions of binary sequences.
Interactive Learning with Pricing for Optimal and Stable Allocations in Markets
Large-scale online recommendation systems must facilitate the allocation of a limited number of items among competing users while learning their preferences from user feedback.
Spectral Regularization Allows Data-frugal Learning over Combinatorial Spaces
Data-driven machine learning models are being increasingly employed in several important inference problems in biology, chemistry, and physics which require learning over combinatorial spaces.
Interactive Recommendations for Optimal Allocations in Markets with Constraints
Recommendation systems when employed in markets play a dual role: they assist users in selecting their most desired items from a large pool and they help in allocating a limited number of items to the users who desire them the most.
Neurotoxin: Durable Backdoors in Federated Learning
In this type of attack, the goal of the attacker is to use poisoned updates to implant so-called backdoors into the learned model such that, at test time, the model's outputs can be fixed to a given target for certain inputs.
Decentralized Competing Bandits in Non-Stationary Matching Markets
We propose and analyze a decentralized and asynchronous learning algorithm, namely Decentralized Non-stationary Competing Bandits (\texttt{DNCB}), where the agents play (restrictive) successive elimination type learning algorithms to learn their preference over the arms.
Minimax Optimal Online Imitation Learning via Replay Estimation
In the tabular setting or with linear function approximation, our meta theorem shows that the performance gap incurred by our approach achieves the optimal $\widetilde{O} \left( \min({H^{3/2}} / {N}, {H} / {\sqrt{N}} \right)$ dependency, under significantly weaker assumptions compared to prior work.
Evaluating natural language processing models with generalization metrics that do not need access to any training or testing data
Our analyses consider (I) hundreds of Transformers trained in different settings, in which we systematically vary the amount of data, the model size and the optimization hyperparameters, (II) a total of 51 pretrained Transformers from eight families of Huggingface NLP models, including GPT2, BERT, etc., and (III) a total of 28 existing and novel generalization metrics.
On the Value of Interaction and Function Approximation in Imitation Learning
In contrast, when the MDP transition structure is known to the learner such as in the case of simulators, we demonstrate fundamental differences compared to the tabular setting in terms of the performance of an optimal algorithm, Mimic-MD (Rajaraman et al. (2020)) when extended to the function approximation setting.
Taxonomizing local versus global structure in neural network loss landscapes
Viewing neural network models in terms of their loss landscapes has a long history in the statistical mechanics approach to learning, and in recent years it has received attention within machine learning proper.
Model Selection for Generic Reinforcement Learning
We propose and analyze a novel algorithm, namely \emph{Adaptive Reinforcement Learning (General)} (\texttt{ARL-GEN}) that adapts to the smallest such family where the true transition kernel $P^*$ lies.
Model Selection for Generic Contextual Bandits
We consider the problem of model selection for the general stochastic contextual bandits under the realizability assumption.
Adaptive Clustering and Personalization in Multi-Agent Stochastic Linear Bandits
We show that, for any agent, the regret scales as $\mathcal{O}(\sqrt{T/N})$, if the agent is in a `well separated' cluster, or scales as $\mathcal{O}(T^{\frac{1}{2} + \varepsilon}/(N)^{\frac{1}{2} -\varepsilon})$ if its cluster is not well separated, where $\varepsilon$ is positive and arbitrarily close to $0$.
LocalNewton: Reducing Communication Bottleneck for Distributed Learning
To enhance practicability, we devise an adaptive scheme to choose L, and we show that this reduces the number of local iterations in worker machines between two model synchronizations as the training proceeds, successively refining the model quality at the master.
Escaping Saddle Points in Distributed Newton's Method with Communication Efficiency and Byzantine Resilience
Moreover, we validate our theoretical findings with experiments using standard datasets and several types of Byzantine attacks, and obtain an improvement of $25\%$ with respect to first order methods in iteration complexity.
Provably Breaking the Quadratic Error Compounding Barrier in Imitation Learning, Optimally
We establish an upper bound $O(|\mathcal{S}|H^{3/2}/N)$ for the suboptimality using the Mimic-MD algorithm in Rajaraman et al (2020) which we prove to be computationally efficient.
BEAR: Sketching BFGS Algorithm for Ultra-High Dimensional Feature Selection in Sublinear Memory
We consider feature selection for applications in machine learning where the dimensionality of the data is so large that it exceeds the working memory of the (local) computing machine.
Training Recommender Systems at Scale: Communication-Efficient Model and Data Parallelism
This is done by identifying and updating only the most relevant neurons of the neural network for each training sample in the data.
FastSecAgg: Scalable Secure Aggregation for Privacy-Preserving Federated Learning
In this paper, we propose a secure aggregation protocol, FastSecAgg, that is efficient in terms of computation and communication, and robust to client dropouts.
Utility-based Resource Allocation and Pricing for Serverless Computing
As one of the fastest-growing cloud services, serverless computing provides an opportunity to better serve both users and providers through the incorporation of market-based strategies for pricing and resource allocation.
Distributed, Parallel, and Cluster Computing Computer Science and Game Theory
Boundary thickness and robustness in learning models
Using these observations, we show that noise-augmentation on mixup training further increases boundary thickness, thereby combating vulnerability to various forms of adversarial attacks and OOD transforms.
An Efficient Framework for Clustered Federated Learning
We address the problem of federated learning (FL) where users are distributed and partitioned into clusters.
Problem-Complexity Adaptive Model Selection for Stochastic Linear Bandits
This is the first algorithm that achieves such model selection guarantees.
Communication-Efficient Gradient Coding for Straggler Mitigation in Distributed Learning
When a particular code is used in this framework, its block-length determines the computation load, dimension determines the communication overhead, and minimum distance determines the straggler tolerance.
Alternating Minimization Converges Super-Linearly for Mixed Linear Regression
Furthermore, we compare AM with a gradient based heuristic algorithm empirically and show that AM dominates in iteration complexity as well as wall-clock time.
Serverless Straggler Mitigation using Local Error-Correcting Codes
Inexpensive cloud services, such as serverless computing, are often vulnerable to straggling nodes that increase end-to-end latency for distributed computation.
Distributed, Parallel, and Cluster Computing Information Theory Information Theory
Communication-Efficient and Byzantine-Robust Distributed Learning with Error Feedback
Moreover, we analyze the compressed gradient descent algorithm with error feedback (proposed in \cite{errorfeed}) in a distributed setting and in the presence of Byzantine worker machines.
SeF: A Secure Fountain Architecture for Slashing Storage Costs in Blockchains
In this paper, we propose an architecture based on 'fountain codes', a class of erasure codes, that enables any full node to 'encode' validated blocks into a small number of 'coded blocks', thereby reducing its storage costs by orders of magnitude.
Cryptography and Security Distributed, Parallel, and Cluster Computing Information Theory Information Theory
Max-Affine Regression: Provable, Tractable, and Near-Optimal Statistical Estimation
Max-affine regression refers to a model where the unknown regression function is modeled as a maximum of $k$ unknown affine functions for a fixed $k \geq 1$.
Robust Federated Learning in a Heterogeneous Environment
Then, leveraging the statistical model, we solve the robust heterogeneous Federated Learning problem \emph{optimally}; in particular our algorithm matches the lower bound on the estimation error in dimension and the number of data points.
Adversarially Trained Autoencoders for Parallel-Data-Free Voice Conversion
Due to the use of a single encoder, our method can generalize to converting the voice of out-of-training speakers to speakers in the training dataset.
Cross-Entropy Loss Leads To Poor Margins
In this work, we study the binary classification of linearly separable datasets and show that linear classifiers could also have decision boundaries that lie close to their training dataset if cross-entropy loss is used for training.
Gradient Coding Based on Block Designs for Mitigating Adversarial Stragglers
In this work, our goal is to construct approximate gradient codes that are resilient to stragglers selected by a computationally unbounded adversary.
OverSketched Newton: Fast Convex Optimization for Serverless Systems
Motivated by recent developments in serverless systems for large-scale computation as well as improvements in scalable randomized matrix algorithms, we develop OverSketched Newton, a randomized Hessian-based optimization algorithm to solve large-scale convex optimization problems in serverless systems.
Cross-Entropy Loss and Low-Rank Features Have Responsibility for Adversarial Examples
We show that differential training can ensure a large margin between the decision boundary of the neural network and the points in the training dataset.
OverSketch: Approximate Matrix Multiplication for the Cloud
We propose OverSketch, an approximate algorithm for distributed matrix multiplication in serverless computing.
Distributed, Parallel, and Cluster Computing Information Theory Information Theory
Greedy Frank-Wolfe Algorithm for Exemplar Selection
In this paper, we consider the problem of selecting representatives from a data set for arbitrary supervised/unsupervised learning tasks.
Rademacher Complexity for Adversarially Robust Generalization
For binary linear classifiers, we prove tight bounds for the adversarial Rademacher complexity, and show that the adversarial Rademacher complexity is never smaller than its natural counterpart, and it has an unavoidable dimension dependence, unless the weight vector has bounded $\ell_1$ norm.
Online Scoring with Delayed Information: A Convex Optimization Viewpoint
We argue that the error in the score estimate accumulated over $T$ iterations is small if the regret of the online convex game is small.
Defending Against Saddle Point Attack in Byzantine-Robust Distributed Learning
In this setting, the Byzantine machines may create fake local minima near a saddle point that is far away from any true local minimum, even when robust gradient estimators are used.
Byzantine-Robust Distributed Learning: Towards Optimal Statistical Rates
In particular, these algorithms are shown to achieve order-optimal statistical error rates for strongly convex losses.
Approximate Ranking from Pairwise Comparisons
Accordingly, we study the problem of finding approximate rankings from pairwise comparisons.
A Sequential Approximation Framework for Coded Distributed Optimization
As an application of the results, we demonstrate solving optimization problems using a sequential approximation approach, which accelerates the algorithm in a distributed system with stragglers.
Gradient Diversity: a Key Ingredient for Scalable Distributed Learning
It has been experimentally observed that distributed implementations of mini-batch stochastic gradient descent (SGD) algorithms exhibit speedup saturation and decaying generalization ability beyond a particular batch-size.
The Sample Complexity of Online One-Class Collaborative Filtering
We consider the online one-class collaborative filtering (CF) problem that consists of recommending items to users over time in an online fashion based on positive ratings only.
Active Ranking from Pairwise Comparisons and when Parametric Assumptions Don't Help
We first analyze a sequential ranking algorithm that counts the number of comparisons won, and uses these counts to decide whether to stop, or to compare another pair of items, chosen based on confidence intervals specified by the data collected up to that point.
CYCLADES: Conflict-free Asynchronous Machine Learning
We present CYCLADES, a general framework for parallelizing stochastic optimization algorithms in a shared memory setting.
Speeding Up Distributed Machine Learning Using Codes
We focus on two of the most basic building blocks of distributed learning algorithms: matrix multiplication and data shuffling.
An Active Learning Framework using Sparse-Graph Codes for Sparse Polynomials and Graph Sketching
By writing the cut function as a polynomial and exploiting the graph structure, we propose a sketching algorithm to learn the an arbitrary $n$-node unknown graph using only few cut queries, which scales {\it almost linearly} in the number of edges and {\it sub-linearly} in the graph size $n$.
Fast and Efficient Sparse 2D Discrete Fourier Transform using Sparse-Graph Codes
For the case when the spatial-domain measurements are corrupted by additive noise, our 2D-FFAST framework extends to a noise-robust version in sub-linear time of O(k log4 N ) using O(k log3 N ) measurements.
Information Theory Multimedia Systems and Control Information Theory
SPRIGHT: A Fast and Robust Framework for Sparse Walsh-Hadamard Transform
We consider the problem of computing the Walsh-Hadamard Transform (WHT) of some $N$-length input vector in the presence of noise, where the $N$-point Walsh spectrum is $K$-sparse with $K = {O}(N^{\delta})$ scaling sub-linearly in the input dimension $N$ for some $0<\delta<1$.
Perturbed Iterate Analysis for Asynchronous Stochastic Optimization
We demonstrate experimentally on a 16-core machine that the sparse and parallel version of SVRG is in some cases more than four orders of magnitude faster than the standard SVRG algorithm.
Parallel Correlation Clustering on Big Graphs
We present C4 and ClusterWild!, two algorithms for parallel correlation clustering that run in a polylogarithmic number of rounds and achieve nearly linear speedups, provably.
Estimation from Pairwise Comparisons: Sharp Minimax Bounds with Topology Dependence
Data in the form of pairwise comparisons arises in many domains, including preference elicitation, sporting competitions, and peer grading among others.
A robust sub-linear time R-FFAST algorithm for computing a sparse DFT
If the DFT X of the signal x has only k non-zero coefficients (where k < n), can we do better?
When is it Better to Compare than to Score?
When eliciting judgements from humans for an unknown quantity, one often has the choice of making direct-scoring (cardinal) or comparative (ordinal) measurements.
