Data deduplication contributes to avoiding great waste of cloud storage resources. Most of current data deduplication schemes focus on file-level data deduplication, without realizing deduplication ownership privacy protection, data blocks retrieval or error data blocks recovery. To this end, this paper proposes cross-clouds encrypted data fine-grained deduplication and recovery (CC-FGDR) scheme supporting blockchain collaborative computing. CC-FGDR exploits two-party joint generation of deduplication tags to prevent deduplication privacy leakage, and constructs vector-commitment ciphertext tree to realize proofs of ownership verification of distributed backup data blocks via blockchain systems. CC-FGDR achieves secure cross-clouds deduplication detection by constructing twin bloom filter tree and privacy puncture pseudo-random function. To achieve data availability, a ciphertext reconstruction matrix is introduced and combined with countable bloom filter to achieve the location, search and retrieval of distributed outsourced data blocks. CC-FGDR further designs low-density parity check code to achieve cross-clouds encrypted error data recovery through collaborative computing of smart contracts. The security analysis demonstrates CC-FGDR ensures data storage verifiability, data confidentiality, and ownership privacy protection. Performance evaluation demonstrates the feasibility of CC-FGDR in the deployment of multiple clouds via blockchain systems, and CC-FGDR significantly improves the deduplication efficiency by joint deduplication in multi-clouds.

Opponent Modeling (OM) aims to enhance decision-making by modeling other agents in multi-agent environments. Existing works typically learn opponent models against a pre-designated fixed set of opponents during training. However, this will cause poor generalization when facing unknown opponents during testing, as previously unseen opponents can exhibit out-of-distribution (OOD) behaviors that the learned opponent models cannot handle. To tackle this problem, we introduce a novel Open-Ended Opponent Modeling (OEOM) framework, which continuously generates opponents with diverse strengths and styles to reduce the possibility of OOD situations occurring during testing. Founded on population-based training and information-theoretic trajectory space diversity regularization, OEOM generates a dynamic set of opponents. This set is then fed to any OM approaches to train a potentially generalizable opponent model. Upon this, we further propose a simple yet effective OM approach that naturally fits within the OEOM framework. This approach is based on in-context reinforcement learning and learns a Transformer that dynamically recognizes and responds to opponents based on their trajectories. Extensive experiments in cooperative, competitive, and mixed environments demonstrate that OEOM is an approach-agnostic framework that improves generalizability compared to training against a fixed set of opponents, regardless of OM approaches or testing opponent settings. The results also indicate that our proposed approach generally outperforms existing OM baselines.

Offline Opponent Modeling (OOM) aims to learn an adaptive autonomous agent policy that dynamically adapts to opponents using an offline dataset from multi-agent games. Previous work assumes that the dataset is optimal. However, this assumption is difficult to satisfy in the real world. When the dataset is suboptimal, existing approaches struggle to work. To tackle this issue, we propose a simple and general algorithmic improvement framework, Truncated Q-driven Instant Policy Refinement (TIPR), to handle the suboptimality of OOM algorithms induced by datasets. The TIPR framework is plug-and-play in nature. Compared to original OOM algorithms, it requires only two extra steps: (1) Learn a horizon-truncated in-context action-value function, namely Truncated Q, using the offline dataset. The Truncated Q estimates the expected return within a fixed, truncated horizon and is conditioned on opponent information. (2) Use the learned Truncated Q to instantly decide whether to perform policy refinement and to generate policy after refinement during testing. Theoretically, we analyze the rationale of Truncated Q from the perspective of No Maximization Bias probability. Empirically, we conduct extensive comparison and ablation experiments in four representative competitive environments. TIPR effectively improves various OOM algorithms pretrained with suboptimal datasets.

Cloud computing has emerged as a promising mode for storaging vast quantities of big data, which is vulnerable to potential security threats, making it urgent to ensure data confidentiality and integrity auditing. In addition, as a large number of duplicate data files exist in cloud storage, data deduplication is significant to improve storage efficiency. In this article, we propose a blockchain-assisted fine-grained deduplication and integrity auditing scheme for outsourced large-scale data in cloud storage, achieving internal deduplication and cross-user external deduplication for ciphertexts and authentication tags. By constructing sparse summation ciphertext tree, the scheme implements Proofs of Ownership protocol through vector commitment, and guarantees the retrieval of distributed deduplication data blocks by designing the reconstruction matrix and bloom filter. The scheme exploits blockchain and smart contracts to ensure transparent integrity auditing without a third-party auditor (TPA), thereby avoiding malicious auditing biases. Security analysis and performance evaluation demonstrate the feasibility of the scheme for deploying in cloud storage systems.

Data auditing contributes to checking the integrity of outsourced data, promoting the vigorous development of cloud storage services. In actual scenarios, such as migration of electronic medical records or data transfer of enterprise mergers and acquisitions, it always require data auditing to help clients with dynamic data migration and integrity checking. In this paper, we present an efficient dynamic certificateless outsourced data auditing mechanism supporting multi-ownership transfer (CDA-MOT), addressing the issue of key escrow and without needing complex certificate management. By integrating a certificateless multi-signature on the same data file into the construction of a homomorphic authenticator based on the Lagrange inverse Multinomial theorem, CDA-MOT not only achieves integrity verification but also enables clients to transfer ownership rights and responsibilities for multi-ownership data in collaboration with cloud servers. Utilizing blockchain systems to store necessary data conversion and update records, as well as smart contracts to fulfill auditing tasks, CDA-MOT owns the characteristics of openness, transparency, accountability, and decentralized public auditing. Besides, CDA-MOT could be further applied in the extension of dynamic update operations, even if outsourced data have been transferred. The security analysis and performance evaluation have demonstrated the feasibility of CDA-MOT in the secure deployment of cloud storage.

Opponent modeling aims at learning the opponent's behaviors, goals, or beliefs to reduce the uncertainty of the competitive environment and assist decision-making. Existing work has mostly focused on learning opponent models online, which is impractical and inefficient in practical scenarios. To this end, we formalize an Offline Opponent Modeling (OOM) problem with the objective of utilizing pre-collected offline datasets to learn opponent models that characterize the opponent from the viewpoint of the controlled agent, which aids in adapting to the unknown fixed policies of the opponent. Drawing on the promises of the Transformers for decision-making, we introduce a general approach, Transformer Against Opponent (TAO), for OOM. Essentially, TAO tackles the problem by harnessing the full potential of the supervised pre-trained Transformers' in-context learning capabilities. The foundation of TAO lies in three stages: an innovative offline policy embedding learning stage, an offline opponent-aware response policy training stage, and a deployment stage for opponent adaptation with in-context learning. Theoretical analysis establishes TAO's equivalence to Bayesian posterior sampling in opponent modeling and guarantees TAO's convergence in opponent policy recognition. Extensive experiments and ablation studies on competitive environments with sparse and dense rewards demonstrate the impressive performance of TAO. Our approach manifests remarkable prowess for fast adaptation, especially in the face of unseen opponent policies, confirming its in-context learning potency.

Opponent modeling is a longstanding research topic aimed at enhancing decision-making by modeling information about opponents in multi-agent environments. However, existing approaches often face challenges such as having difficulty generalizing to unknown opponent policies and conducting unstable performance. To tackle these challenges, we propose a novel approach based on in-context learning and decision-time search named Opponent Modeling with In-context Search (OMIS). OMIS leverages in-context learning-based pretraining to train a Transformer model for decision-making. It consists of three in-context components: an actor learning best responses to opponent policies, an opponent imitator mimicking opponent actions, and a critic estimating state values. When testing in an environment that features unknown non-stationary opponent agents, OMIS uses pretrained in-context components for decision-time search to refine the actor's policy. Theoretically, we prove that under reasonable assumptions, OMIS without search converges in opponent policy recognition and has good generalization properties; with search, OMIS provides improvement guarantees, exhibiting performance stability. Empirically, in competitive, cooperative, and mixed environments, OMIS demonstrates more effective and stable adaptation to opponents than other approaches. See our project website at https://sites.google.com/view/nips2024-omis.