Privacy-preserving Transformer inference has gained attention due to the potential leakage of private information. Despite recent progress, existing frameworks still fall short of practical model scales, with gaps up to a hundredfold. A possible way to close this gap is the Mixture of Experts (MoE) architecture, which has emerged as a promising technique to scale up model capacity with minimal overhead. However, given that the current secure two-party (2-PC) protocols allow the server to homomorphically compute the FFN layer with its plaintext model weight, under the MoE setting, this could reveal which expert is activated to the server, exposing token-level privacy about the client's input. While naively evaluating all the experts before selection could protect privacy, it nullifies MoE sparsity and incurs the heavy computational overhead that sparse MoE seeks to avoid. To address the privacy and efficiency limitations above, we propose a 2-PC privacy-preserving inference framework, SecMoE. Unifying per-entry circuits in both the MoE layer and piecewise polynomial functions, SecMoE obliviously selects the extracted parameters from circuits and only computes one encrypted entry, which we refer to as Select-Then-Compute. This makes the model for private inference scale to 63× larger while only having a 15.2× increase in end-to-end runtime. Extensive experiments show that, under 5 expert settings, SecMoE lowers the end-to-end private inference communication by 1.8~7.1× and achieves 1.3~3.8× speedup compared to the state-of-the-art (SOTA) protocols.

Secure two-party computation (2PC) enables two parties to jointly evaluate a function while maintaining input privacy. Despite recent significant progress, a notable efficiency gap remains between actively secure and passively secure protocols. In S&P'12, Huang, Katz, and Evans formalized the notion of active security with one-bit leakage, providing a promising approach to bridging this gap. Protocols derived from this notion have become foundational in designing highly efficient actively secure 2PC protocols. However, a critical challenge identified by Huang, Katz, and Evans remains unexplored: these protocols face significant weaknesses in ensuring fairness for honest parties when employed in standalone settings rather than as components within larger protocols. While the authors proposed two potential solutions to mitigate this issue, both approaches are prohibitively expensive and lack formalization of security guarantees. In this paper, we first formally define an enhanced notion called active security with one-bit-advantage bound, in which the adversaries' advantages are strictly bounded to at most one bit beyond what honest parties obtain. This bound is enforced through a progressive revelation mechanism, where the evaluation result is disclosed incrementally bit by bit. In addition, we propose a novel approach leveraging label structures within garbled circuits to design a highly efficient constant-round 2PC protocol that achieves active security with one-bit advantage bound. Our protocol demonstrates runtime performance nearly identical to that of passively secure garbled-circuit counterparts in duplex networks (e.g., 1.033 × for the SHA256 circuit in LAN), with low overhead for output progressive revelation (only 80 communicated bytes per bit release). With its strengthened security guarantees and minimal overhead, our protocol is highly suitable for practical 2PC applications.

The widespread adoption of smartphones and Location-Based Social Networks has led to a massive influx of spatio-temporal data, creating unparalleled opportunities for enhancing Point-of-Interest (POI) recommendation systems. These advanced POI systems are crucial for enriching user experiences, enabling personalized interactions, and optimizing decision-making processes in the digital landscape. However, existing surveys tend to focus on traditional approaches and few of them delve into cutting-edge developments, emerging architectures, as well as security considerations in POI recommendations. To address this gap, our survey stands out by offering a comprehensive, up-to-date review of POI recommendation systems, covering advancements in models, architectures, and security aspects. We systematically examine the transition from traditional models to advanced techniques such as large language models. Additionally, we explore the architectural evolution from centralized to decentralized and federated learning systems, highlighting the improvements in scalability and privacy. Furthermore, we address the increasing importance of security, examining potential vulnerabilities and privacy-preserving approaches. Our taxonomy provides a structured overview of the current state of POI recommendation, while we also identify promising directions for future research in this rapidly advancing field.