Drawing upon recent advances in language model alignment, we formulate offline Reinforcement Learning as a two-stage optimization problem: First pretraining expressive generative policies on reward-free behavior datasets, then finetuning these policies to align with task-specific annotations like Q-values. This strategy allows us to leverage abundant and diverse behavior data to enhance generalization and enable rapid adaptation to downstream tasks using minimal annotations. In particular, we introduce Efficient Diffusion Alignment (EDA) for solving continuous control problems. EDA utilizes diffusion models for behavior modeling. However, unlike previous approaches, we represent diffusion policies as the derivative of a scalar neural network with respect to action inputs. This representation is critical because it enables direct density calculation for diffusion models, making them compatible with existing LLM alignment theories. During policy fine-tuning, we extend preference-based alignment methods like Direct Preference Optimization (DPO) to align diffusion behaviors with continuous Q-functions. Our evaluation on the D4RL benchmark shows that EDA exceeds all baseline methods in overall performance. Notably, EDA maintains about 95\% of performance and still outperforms several baselines given only 1\% of Q-labelled data during fine-tuning.
论文检索
输入标题、作者或关键词，从 14,273 篇学术成果中精准定位
Multimodal Models
Training models with longer in-context lengths is a significant challenge for multimodal machine learning due to substantial GPU memory and computational costs. This exploratory study does not present state-of-the-art models; rather, it introduces an innovative method designed to increase in-context text length in multi-modality large language models (MLLMs) efficiently. We present \ModelFullName (\ModelName), which processes long in-context text using visual tokens. This technique significantly reduces GPU memory usage and floating point operations (FLOPs). For instance, our method expands the pre-training in-context length from 256 to 2048 tokens with fewer FLOPs for a 56 billion parameter MOE model. Experimental results demonstrate that \ModelName enhances OCR capabilities and delivers superior performance on common downstream benchmarks for in-context few-shot evaluation. Additionally, \ModelName proves effective for long context inference, achieving results comparable to full text input while maintaining computational efficiency.
Optimization · Optimization and Learning under Uncertainty
Uncertain perturbations in dynamical systems often arise from diverse resources, represented by latent components. The predictions for these components, typically generated by "black-box" machine learning tools, are prone to inaccuracies. To tackle this challenge, we introduce DISC, a novel policy that learns a confidence parameter online to harness the potential of accurate predictions while also mitigating the impact of erroneous forecasts. When predictions are precise, DISC leverages this information to achieve near-optimal performance. Conversely, in the case of significant prediction errors, it still has a worst-case competitive ratio guarantee. We provide competitive ratio bounds for DISC under both linear mixing of latent variables as well as a broader class of mixing functions. Our results highlight a first-of-its-kind "best-of-both-worlds" integration of machine-learned predictions, thus lead to a near-optimal consistency and robustness tradeoff, which provably improves what can be obtained without learning the confidence parameter. We validate the applicability of DISC across a spectrum of practical scenarios.
Theory · Reinforcement Learning and Planning
Hybrid Reinforcement Learning (RL), where an agent learns from both an offline dataset and online explorations in an unknown environment, has garnered significant recent interest. A crucial question posed by Xie et al. (2022) is whether hybrid RL can improve upon the existing lower bounds established in purely offline and purely online RL without relying on the single-policy concentrability assumption. While Li et al. (2023) provided an affirmative answer to this question in the tabular PAC RL case, the question remains unsettled for both the regret-minimizing RL case and the non-tabular case. In this work, building upon recent advancements in offline RL and reward-agnostic exploration, we develop computationally efficient algorithms for both PAC and regret-minimizing RL with linear function approximation, without requiring concentrability on the entire state-action space. We demonstrate that these algorithms achieve sharper error or regret bounds that are no worse than, and can improve on, the optimal sample complexity in offline RL (the first algorithm, for PAC RL) and online RL (the second algorithm, for regret-minimizing RL) in linear Markov decision processes (MDPs), regardless of the quality of the behavior policy. To our knowledge, this work establishes the tightest theoretical guarantees currently available for hybrid RL in linear MDPs.
We propose a novel approach for training large language models (LLMs) to adhere to objectives defined within a latent embedding space. Our method leverages reinforcement learning (RL), treating a pre-trained LLM as an environment. Our embedding-aligned guided language (EAGLE) agent is trained to iteratively steer the LLM's generation towards optimal regions of the latent embedding space, w.r.t. some predefined criterion. We demonstrate the effectiveness of the EAGLE agent using the MovieLens 25M and Amazon Review datasets to surface content gaps that satisfy latent user demand. We also demonstrate the benefit of using an optimal design of a state-dependent action set to improve EAGLE's efficiency. Our work paves the way for controlled and grounded text generation using LLMs, ensuring consistency with domain-specific knowledge and data representations.
Reinforcement Learning · Deep RL
In safety-critical RL settings, the inclusion of an additional cost function is often favoured over the arduous task of modifying the reward function to ensure the agent's safe behaviour. However, designing or evaluating such a cost function can be prohibitively expensive. For instance, in the domain of self-driving, designing a cost function that encompasses all unsafe behaviours (e.g., aggressive lane changes, risky overtakes) is inherently complex, it must also consider all the actors present in the scene making it expensive to evaluate. In such scenarios, the cost function can be learned from feedback collected offline in between training rounds. This feedback can be system generated or elicited from a human observing the training process. Previous approaches have not been able to scale to complex environments and are constrained to receiving feedback at the state level which can be expensive to collect. To this end, we introduce an approach that scales to more complex domains and extends beyond state-level feedback, thus, reducing the burden on the evaluator. Inferring the cost function in such settings poses challenges, particularly in assigning credit to individual states based on trajectory-level feedback. To address this, we propose a surrogate objective that transforms the problem into a state-level supervised classification task with noisy labels, which can be solved efficiently. Additionally, it is often infeasible to collect feedback for every trajectory generated by the agent, hence, two fundamental questions arise: (1) Which trajectories should be presented to the human? and (2) How many trajectories are necessary for effective learning? To address these questions, we introduce a \textit{novelty-based sampling} mechanism that selectively involves the evaluator only when the the agent encounters a \textit{novel} trajectory, and discontinues querying once the trajectories are no longer \textit{novel}. We showcase the efficiency of our method through experimentation on several benchmark Safety Gymnasium environments and realistic self-driving scenarios. Our method demonstrates near-optimal performance, comparable to when the cost function is known, by relying solely on trajectory-level feedback across multiple domains. This highlights both the effectiveness and scalability of our approach. The code to replicate these results can be found at \href{https://github.com/shshnkreddy/RLSF}{https://github.com/shshnkreddy/RLSF}
Optimization · Optimization and Learning under Uncertainty
We study a general clustering setting in which we have $n$ elements to be clustered, and we aim to perform as few queries as possible to an oracle that returns a noisy sample of the weighted similarity between two elements. Our setting encompasses many application domains in which the similarity function is costly to compute and inherently noisy. We introduce two novel formulations of online learning problems rooted in the paradigm of Pure Exploration in Combinatorial Multi-Armed Bandits (PE-CMAB): fixed confidence and fixed budget settings. For both settings, we design algorithms that combine a sampling strategy with a classic approximation algorithm for correlation clustering and study their theoretical guarantees. Our results are the first examples of polynomial-time algorithms that work for the case of PE-CMAB in which the underlying offline optimization problem is NP-hard.
Probabilistic Methods
Estimating and disentangling epistemic uncertainty, uncertainty that is reducible with more training data, and aleatoric uncertainty, uncertainty that is inherent to the task at hand, is critically important when applying machine learning to high-stakes applications such as medical imaging and weather forecasting. Conditional diffusion models' breakthrough ability to accurately and efficiently sample from the posterior distribution of a dataset now makes uncertainty estimation conceptually straightforward: One need only train and sample from a large ensemble of diffusion models. Unfortunately, training such an ensemble becomes computationally intractable as the complexity of the model architecture grows. In this work we introduce a new approach to ensembling, hyper-diffusion models (HyperDM), which allows one to accurately estimate both epistemic and aleatoric uncertainty with a single model. Unlike existing single-model uncertainty methods like Monte-Carlo dropout and Bayesian neural networks, HyperDM offers prediction accuracy on par with, and in some cases superior to, multi-model ensembles. Furthermore, our proposed approach scales to modern network architectures such as Attention U-Net and yields more accurate uncertainty estimates compared to existing methods. We validate our method on two distinct real-world tasks: x-ray computed tomography reconstruction and weather temperature forecasting.
Pre-trained vision-language models like CLIP have remarkably adapted to various downstream tasks. Nonetheless, their performance heavily depends on the specificity of the input text prompts, which requires skillful prompt template engineering. Instead, current approaches to prompt optimization learn the prompts through gradient descent, where the prompts are treated as adjustable parameters. However, these methods tend to lead to overfitting of the base classes seen during training and produce prompts that are no longer understandable by humans. This paper introduces a simple but interpretable prompt optimizer (IPO), that utilizes large language models (LLMs) to generate textual prompts dynamically. We introduce a Prompt Optimization Prompt that not only guides LLMs in creating effective prompts but also stores past prompts with their performance metrics, providing rich in-context information. Additionally, we incorporate a large multimodal model (LMM) to condition on visual content by generating image descriptions, which enhance the interaction between textual and visual modalities. This allows for the creation of dataset-specific prompts that improve generalization performance, while maintaining human comprehension. Extensive testing across 11 datasets reveals that IPO not only improves the accuracy of existing gradient-descent-based prompt learning methods but also considerably enhances the interpretability of the generated prompts. By leveraging the strengths of LLMs, our approach ensures that the prompts remain human-understandable, thereby facilitating better transparency and oversight for vision-language models.
Deep Learning · Attention Mechanisms
Large Language Models (LLMs) have the capacity to store and recall facts. Through experimentation with open-source models, we observe that this ability to retrieve facts can be easily manipulated by changing contexts, even without altering their factual meanings. These findings highlight that LLMs might behave like an associative memory model where certain tokens in the contexts serve as clues to retrieving facts. We mathematically explore this property by studying how transformers, the building blocks of LLMs, can complete such memory tasks. We study a simple latent concept association problem with a one-layer transformer and we show theoretically and empirically that the transformer gathers information using self-attention and uses the value matrix for associative memory.
Interpretability
The lack of transparency in the decision-making processes of deep learning systems presents a significant challenge in modern artificial intelligence (AI), as it impairs users’ ability to rely on and verify these systems. To address this challenge, Concept Bottleneck Models (CBMs) have made significant progress by incorporating human-interpretable concepts into deep learning architectures. This approach allows predictions to be traced back to specific concept patterns that users can understand and potentially intervene on. However, existing CBMs’ task predictors are not fully interpretable, preventing a thorough analysis and any form of formal verification of their decision-making process prior to deployment, thereby raising significant reliability concerns. To bridge this gap, we introduce Concept-based Memory Reasoner (CMR), a novel CBM designed to provide a human-understandable and provably-verifiable task prediction process. Our approach is to model each task prediction as a neural selection mechanism over a memory of learnable logic rules, followed by a symbolic evaluation of the selected rule. The presence of an explicit memory and the symbolic evaluation allow domain experts to inspect and formally verify the validity of certain global properties of interest for the task prediction process. Experimental results demonstrate that CMR achieves better accuracy-interpretability trade-offs to state-of-the-art CBMs, discovers logic rules consistent with ground truths, allows for rule interventions, and allows pre-deployment verification.
Time Series
*Boundaries* are the timestamps at which a class in a time series changes. Recently, representation-based boundary detection has gained popularity, but its emphasis on consecutive distance difference backfires, especially when the changes are gradual. In this paper, we propose a boundary detection method, **RECURVE**, based on a novel change metric, the ***curvature*** of a representation trajectory, to accommodate both gradual and abrupt changes. Here, a sequence of representations in the representation space is interpreted as a trajectory, and a curvature at each timestamp can be computed. Using the theory of random walk, we formally show that the mean curvature is lower near boundaries than at other points. Extensive experiments using diverse real-world time-series datasets confirm the superiority of RECURVE over state-of-the-art methods.
Reinforcement Learning · Deep RL
Reinforcement learning (RL) algorithms have been very successful at tackling complex control problems, such as AlphaGo or fusion control. However, current research mainly emphasizes solution quality, often achieved by using large models trained on large amounts of data, and does not account for the financial, environmental, and societal costs associated with developing and deploying such models. Modern neural networks are often overparameterized and a significant number of parameters can be pruned without meaningful loss in performance, resulting in more efficient use of the model's capacity lottery ticket. We present a methodology for identifying sub-networks within a larger network in reinforcement learning (RL). We call such sub-networks, neural pathways. We show empirically that even very small learned sub-networks, using less than 5% of the large network's parameters, can provide very good quality solutions. We also demonstrate the training of multiple pathways within the same networks in a multitask setup, where each pathway is encouraged to tackle a separate task. We evaluate empirically our approach on several continuous control tasks, in both online and offline training
Reinforcement Learning · Deep RL
Decision Transformer (DT), as one of the representative Reinforcement Learning via Supervised Learning (RvS) methods, has achieved strong performance in offline learning tasks by leveraging the powerful Transformer architecture for sequential decision-making. However, in adversarial environments, these methods can be non-robust, since the return is dependent on the strategies of both the decision-maker and adversary. Training a probabilistic model conditioned on observed return to predict action can fail to generalize, as the trajectories that achieve a return in the dataset might have done so due to a suboptimal behavior adversary. To address this, we propose a worst-case-aware RvS algorithm, the Adversarially Robust Decision Transformer (ARDT), which learns and conditions the policy on in-sample minimax returns-to-go. ARDT aligns the target return with the worst-case return learned through minimax expectile regression, thereby enhancing robustness against powerful test-time adversaries. In experiments conducted on sequential games with full data coverage, ARDT can generate a maximin (Nash Equilibrium) strategy, the solution with the largest adversarial robustness. In large-scale sequential games and continuous adversarial RL environments with partial data coverage, ARDT demonstrates significantly superior robustness to powerful test-time adversaries and attains higher worst-case returns compared to contemporary DT methods.
Learning Theory
Transformers have achieved great success in recent years. Interestingly, transformers have shown particularly strong in-context learning capability -- even without fine-tuning, they are still able to solve unseen tasks well purely based on task-specific prompts. In this paper, we study the capability of one-layer transformers in learning the one-nearest neighbor prediction rule. Under a theoretical framework where the prompt contains a sequence of labeled training data and unlabeled test data, we show that, although the loss function is nonconvex, when trained with gradient descent, a single softmax attention layer can successfully learn to behave like a one-nearest neighbor classifier. Our result gives a concrete example on how transformers can be trained to implement nonparametric machine learning algorithms, and sheds light on the role of softmax attention in transformer models.
Trustworthy Machine Learning
Federated learning (FL) is an appealing paradigm that allows a group of machines(a.k.a. clients) to learn collectively while keeping their data local. However, dueto the heterogeneity between the clients’ data distributions, the model obtainedthrough the use of FL algorithms may perform poorly on some client’s data.Personalization addresses this issue by enabling each client to have a differentmodel tailored to their own data while simultaneously benefiting from the otherclients’ data. We consider an FL setting where some clients can be adversarial, andwe derive conditions under which full collaboration fails. Specifically, we analyzethe generalization performance of an interpolated personalized FL framework in thepresence of adversarial clients, and we precisely characterize situations when fullcollaboration performs strictly worse than fine-tuned personalization. Our analysisdetermines how much we should scale down the level of collaboration, accordingto data heterogeneity and the tolerable fraction of adversarial clients. We supportour findings with empirical results on mean estimation and binary classificationproblems, considering synthetic and benchmark image classification datasets
The deep operator networks (DeepONet), a class of neural operators that learn mappings between function spaces, have recently been developed as surrogate models for parametric partial differential equations (PDEs). In this work we propose a derivative-enhanced deep operator network (DE-DeepONet), which leverages derivative information to enhance the solution prediction accuracy and provides a more accurate approximation of solution-to-parameter derivatives, especially when training data are limited. DE-DeepONet explicitly incorporates linear dimension reduction of high dimensional parameter input into DeepONet to reduce training cost and adds derivative loss in the loss function to reduce the number of required parameter-solution pairs. We further demonstrate that the use of derivative loss can be extended to enhance other neural operators, such as the Fourier neural operator (FNO). Numerical experiments validate the effectiveness of our approach.
Deep Learning · Robustness
We propose RABBit, a Branch-and-Bound-based verifier for verifying relational properties defined over Deep Neural Networks, such as robustness against universal adversarial perturbations (UAP). Existing SOTA complete $L_{\infty}$-robustness verifiers can not reason about dependencies between multiple executions and, as a result, are imprecise for relational verification. In contrast, existing SOTA relational verifiers only apply a single bounding step and do not utilize any branching strategies to refine the obtained bounds, thus producing imprecise results. We develop the first scalable Branch-and-Bound-based relational verifier, RABBit, which efficiently combines branching over multiple executions with cross-executional bound refinement to utilize relational constraints, gaining substantial precision over SOTA baselines on a wide range of datasets and networks. Our code is at https://github.com/uiuc-focal-lab/RABBit.
Reinforcement Learning · Everything Else
The optimal policy in various real-world strategic decision-making problems depends both on the environmental configuration and exogenous events. For these settings, we introduce Contextual Bilevel Reinforcement Learning (CB-RL), a stochastic bilevel decision-making model, where the lower level consists of solving a contextual Markov Decision Process (CMDP). CB-RL can be viewed as a Stackelberg Game where the leader and a random context beyond the leader’s control together decide the setup of many MDPs that potentially multiple followers best respond to. This framework extends beyond traditional bilevel optimization and finds relevance in diverse fields such as RLHF, tax design, reward shaping, contract theory and mechanism design. We propose a stochastic Hyper Policy Gradient Descent (HPGD) algorithm to solve CB-RL, and demonstrate its convergence. Notably, HPGD uses stochastic hypergradient estimates, based on observations of the followers’ trajectories. Therefore, it allows followers to use any training procedure and the leader to be agnostic of the specific algorithm, which aligns with various real-world scenarios. We further consider the setting when the leader can influence the training of followers and propose an accelerated algorithm. We empirically demonstrate the performance of our algorithm for reward shaping and tax design.
Recently, 3D Gaussian Splatting (3DGS) has become a promising framework for novel view synthesis, offering fast rendering speeds and high fidelity. However, the large number of Gaussians and their associated attributes require effective compression techniques. Existing methods primarily compress neural Gaussians individually and independently, i.e., coding all the neural Gaussians at the same time, with little design for their interactions and spatial dependence. Inspired by the effectiveness of the context model in image compression, we propose the first autoregressive model at the anchor level for 3DGS compression in this work. We divide anchors into different levels and the anchors that are not coded yet can be predicted based on the already coded ones in all the coarser levels, leading to more accurate modeling and higher coding efficiency. To further improve the efficiency of entropy coding, e.g., to code the coarsest level with no already coded anchors, we propose to introduce a low-dimensional quantized feature as the hyperprior for each anchor, which can be effectively compressed. Our work pioneers the context model in the anchor level for 3DGS representation, yielding an impressive size reduction of over 100 times compared to vanilla 3DGS and 15 times compared to the most recent state-of-the-art work Scaffold-GS, while achieving comparable or even higher rendering quality.