What Lurks Within? Concept Auditing for Shared Diffusion Models at Scale

Diffusion models (DMs) have revolutionized text-to-image generation, enabling the creation of highly realistic and customized images from text prompts. With the rise of parameter-efficient fine-tuning (PEFT) techniques like LoRA, users can now customize powerful pre-trained models using minimal computational resources. However, the widespread sharing of fine-tuned DMs on open platforms raises growing ethical and legal concerns, as these models may inadvertently or deliberately generate sensitive or unauthorized content, such as copyrighted material, private individuals, or harmful content. Despite the increasing regulatory attention on generative AI, there are currently no practical tools for systematically auditing these models before deployment. In this paper, we address the problem of concept auditing: determining whether a fine-tuned DM has learned to generate a specific target concept. Existing approaches typically rely on prompt-based input crafting and output-based image classification but suffer from critical limitations, including prompt uncertainty, concept drift, and poor scalability. To overcome these challenges, we introduce Prompt-Agnostic Image-Free Auditing (PAIA), a novel, model-centric concept auditing framework. By treating the DM as the object of inspection, PAIA enables direct analysis of internal model behavior, bypassing the need for optimized prompts or generated images. We evaluate PAIA on 320 controlled model and 690 real-world community models sourced from a public DM sharing platform. PAIA achieves over 90% detection accuracy while reducing auditing time by 18-40x compared to existing baselines. To our knowledge, PAIA is the first scalable and practical solution for pre-deployment concept auditing of diffusion models, providing a practical foundation for safer and more transparent diffusion model sharing.

You Don't Need All Attentions: Distributed Dynamic Fine-Tuning for Foundation Models

Fine-tuning plays a crucial role in adapting models to downstream tasks with minimal training efforts. However, the rapidly increasing size of foundation models poses a daunting challenge for accommodating foundation model fine-tuning in most commercial devices, which often have limited memory bandwidth. Techniques like model sharding and tensor parallelism address this issue by distributing computation across multiple devices to meet memory requirements. Nevertheless, these methods do not fully leverage their foundation nature in facilitating the fine-tuning process, resulting in high computational costs and imbalanced workloads. We introduce a novel Distributed Dynamic Fine-Tuning (D2FT) framework that strategically orchestrates operations across attention modules based on our observation that not all attention modules are necessary for forward and backward propagation in fine-tuning foundation models. Through three innovative selection strategies, D2FT significantly reduces the computational workload required for fine-tuning foundation models. Furthermore, D2FT addresses workload imbalances in distributed computing environments by optimizing these selection strategies via multiple knapsack optimization. Our experimental results demonstrate that the proposed D2FT framework reduces the training computational costs by 40% and training communication costs by 50% with only 1% to 2% accuracy drops on the CIFAR-10, CIFAR-100, and Stanford Cars datasets. Moreover, the results show that D2FT can be effectively extended to recent LoRA, a state-of-the-art parameter-efficient fine-tuning technique. By reducing 40% computational cost or 50% communication cost, D2FT LoRA top-1 accuracy only drops 4% to 6% on Stanford Cars dataset.

A semantic communication-based workload-adjustable transceiver for wireless AI-generated content (AIGC) delivery

With the significant advances in generative AI (GAI) and the proliferation of mobile devices, providing high-quality AI-generated content (AIGC) services via wireless networks is becoming the future direction. However, the primary challenges of AIGC service delivery in wireless networks lie in unstable channels, limited bandwidth resources, and unevenly distributed computational resources. In this paper, we employ semantic communication (SemCom) in diffusion-based GAI models to propose a Resource-aware wOrkload-adjUstable TransceivEr (ROUTE) for AIGC delivery in dynamic wireless networks. Specifically, to relieve the communication resource bottleneck, SemCom is utilized to prioritize semantic information of the generated content. Then, to improve computational resource utilization in both edge and local and reduce AIGC semantic distortion in transmission, modified diffusion-based models are applied to adjust the computing workload and semantic density in cooperative content generation. Simulations verify the superiority of our proposed ROUTE in terms of latency and content quality compared to conventional AIGC approaches.

Formalizing Complex Mathematical Statements with LLMs: A Study on Mathematical Definitions

Thanks to their linguistic capabilities, LLMs offer an opportunity to bridge the gap between informal mathematics and formal languages through autoformalization. However, it is still unclear how well LLMs generalize to sophisticated and naturally occurring mathematical statements. To address this gap, we investigate the task of autoformalizing real-world mathematical definitions -- a critical component of mathematical discourse. Specifically, we introduce two novel resources for autoformalisation, collecting definitions from Wikipedia (Def_Wiki) and arXiv papers (Def_ArXiv). We then systematically evaluate a range of LLMs, analyzing their ability to formalize definitions into Isabelle/HOL. Furthermore, we investigate strategies to enhance LLMs' performance including refinement through external feedback from Proof Assistants, and formal definition grounding, where we guide LLMs through relevant contextual elements from formal mathematical libraries. Our findings reveal that definitions present a greater challenge compared to existing benchmarks, such as miniF2F. In particular, we found that LLMs still struggle with self-correction, and aligning with relevant mathematical libraries. At the same time, structured refinement methods and definition grounding strategies yield notable improvements of up to 16% on self-correction capabilities and 43% on the reduction of undefined errors, highlighting promising directions for enhancing LLM-based autoformalization in real-world scenarios.

Power-Efficient Optimization for Coexisting Semantic and Bit-Based Users in NOMA Networks

Semantic communication focuses on transmitting the meaning of data, aiming for efficient, relevant communication, while non-orthogonal multiple access (NOMA) enhances spectral efficiency by allowing multiple users to share the same spectrum. Integrating semantic users into a NOMA network with bit-based users improves both transmission and spectrum efficiency. However, the performance metric for semantic communication differs significantly from that of traditional communication, posing challenges in simultaneously meeting individual user demands and minimizing transmission power, especially in scenarios with coexisting semantic and bit-based users. Furthermore, the different hardware architectures of semantic and bit-based users complicate the implementation of successive interference cancellation (SIC). To address these challenges, in this paper, we propose a clustered framework to mitigate the complexity of SIC and two multiple access (MA) schemes, e.g., pure cluster-based NOMA (P-CNOMA) and hybrid cluster-based NOMA (H-CNOMA), to minimize the total transmission power. The P-CNOMA scheme can achieve the minimum transmission power, but may not satisfy the high quality of service (QoS) requirement. In contrast, H-CNOMA addresses these issues with a slight increase in power and a reduced semantic rate. These two schemes complement each other, enabling an adaptive MA selection mechanism that adapts to specific network conditions and user requirements.

RemoteRAG: A Privacy-Preserving LLM Cloud RAG Service

Retrieval-augmented generation (RAG) improves the service quality of large language models by retrieving relevant documents from credible literature and integrating them into the context of the user query. Recently, the rise of the cloud RAG service has made it possible for users to query relevant documents conveniently. However, directly sending queries to the cloud brings potential privacy leakage. In this paper, we are the first to formally define the privacy-preserving cloud RAG service to protect the user query and propose RemoteRAG as a solution regarding privacy, efficiency, and accuracy. For privacy, we introduce $(n,\epsilon)$-DistanceDP to characterize privacy leakage of the user query and the leakage inferred from relevant documents. For efficiency, we limit the search range from the total documents to a small number of selected documents related to a perturbed embedding generated from $(n,\epsilon)$-DistanceDP, so that computation and communication costs required for privacy protection significantly decrease. For accuracy, we ensure that the small range includes target documents related to the user query with detailed theoretical analysis. Experimental results also demonstrate that RemoteRAG can resist existing embedding inversion attack methods while achieving no loss in retrieval under various settings. Moreover, RemoteRAG is efficient, incurring only $0.67$ seconds and $46.66$KB of data transmission ($2.72$ hours and $1.43$ GB with the non-optimized privacy-preserving scheme) when retrieving from a total of $10^6$ documents.

DeepCore: Simple Fingerprint Construction for Differentiating Homologous and Piracy Models

As intellectual property rights, the copyright protection of deep models is becoming increasingly important. Existing work has made many attempts at model watermarking and fingerprinting, but they have ignored homologous models trained with similar structures or training datasets. We highlight challenges in efficiently querying black-box piracy models to protect model copyrights without misidentifying homologous models. To address these challenges, we propose a novel method called DeepCore, which discovers that the classification confidence of the model is positively correlated with the distance of the predicted sample from the model decision boundary and piracy models behave more similarly at high-confidence classified sample points. Then DeepCore constructs core points far away from the decision boundary by optimizing the predicted confidence of a few sample points and leverages behavioral discrepancies between piracy and homologous models to identify piracy models. Finally, we design different model identification methods, including two similarity-based methods and a clustering-based method to identify piracy models using models' predictions of core points. Extensive experiments show the effectiveness of DeepCore in identifying various piracy models, achieving lower missed and false identification rates, and outperforming state-of-the-art methods.

Consistent Autoformalization for Constructing Mathematical Libraries

Autoformalization is the task of automatically translating mathematical content written in natural language to a formal language expression. The growing language interpretation capabilities of Large Language Models (LLMs), including in formal languages, are lowering the barriers for autoformalization. However, LLMs alone are not capable of consistently and reliably delivering autoformalization, in particular as the complexity and specialization of the target domain grows. As the field evolves into the direction of systematically applying autoformalization towards large mathematical libraries, the need to improve syntactic, terminological and semantic control increases. This paper proposes the coordinated use of three mechanisms, most-similar retrieval augmented generation (MS-RAG), denoising steps, and auto-correction with syntax error feedback (Auto-SEF) to improve autoformalization quality. The empirical analysis, across different models, demonstrates that these mechanisms can deliver autoformalizaton results which are syntactically, terminologically and semantically more consistent. These mechanisms can be applied across different LLMs and have shown to deliver improve results across different model types.

Safety Layers of Aligned Large Language Models: The Key to LLM Security

Aligned LLMs are highly secure, capable of recognizing and refusing to answer malicious questions. However, the role of internal parameters in maintaining this security is not well understood, further these models are vulnerable to security degradation when fine-tuned with non-malicious backdoor data or normal data. To address these challenges, our work uncovers the mechanism behind security in aligned LLMs at the parameter level, identifying a small set of contiguous layers in the middle of the model that are crucial for distinguishing malicious queries from normal ones, referred to as "safety layers." We first confirm the existence of these safety layers by analyzing variations in input vectors within the model's internal layers. Additionally, we leverage the over-rejection phenomenon and parameters scaling analysis to precisely locate the safety layers. Building on this understanding, we propose a novel fine-tuning approach, Safely Partial-Parameter Fine-Tuning (SPPFT), that fixes the gradient of the safety layers during fine-tuning to address the security degradation. Our experiments demonstrate that this approach significantly preserves model security while maintaining performance and reducing computational resources compared to full fine-tuning.

Hide Your Malicious Goal Into Benign Narratives: Jailbreak Large Language Models through Neural Carrier Articles

Jailbreak attacks on Language Model Models (LLMs) entail crafting prompts aimed at exploiting the models to generate malicious content. This paper proposes a new type of jailbreak attacks which shift the attention of the LLM by inserting a prohibited query into a carrier article. The proposed attack leverage the knowledge graph and a composer LLM to automatically generating a carrier article that is similar to the topic of the prohibited query but does not violate LLM's safeguards. By inserting the malicious query to the carrier article, the assembled attack payload can successfully jailbreak LLM. To evaluate the effectiveness of our method, we leverage 4 popular categories of ``harmful behaviors'' adopted by related researches to attack 6 popular LLMs. Our experiment results show that the proposed attacking method can successfully jailbreak all the target LLMs which high success rate, except for Claude-3.