Text-to-image diffusion models have emerged as powerful tools for generating high-quality images from textual descriptions. However, their increasing popularity has raised significant copyright concerns, as these models can be misused to reproduce copyrighted content without authorization. In response, recent studies have proposed various copyright protection methods, including adversarial perturbation, concept erasure, and watermarking techniques. However, their effectiveness and robustness against advanced attacks remain largely unexplored. Moreover, the lack of unified evaluation frameworks has hindered systematic comparison and fair assessment of different approaches. To bridge this gap, we systematize existing copyright protection methods and attacks, providing a unified taxonomy of their design spaces. We then develop CopyrightMeter, a unified evaluation framework that incorporates 17 state-of-the-art protections and 16 representative attacks. Leveraging CopyrightMeter, we comprehensively evaluate protection methods across multiple dimensions, thereby uncovering how different design choices impact fidelity, efficacy, and resilience under attacks. Our analysis reveals several key findings: (i) most protections (16/17) are not resilient against attacks; (ii) the "best" protection varies depending on the target priority; (iii) more advanced attacks significantly promote the upgrading of protections. These insights provide concrete guidance for developing more robust protection methods, while its unified evaluation protocol establishes a standard benchmark for future copyright protection research in text-to-image generation.

Advancements in large language models (LLMs) have renewed concerns about AI alignment - the consistency between human and AI goals and values. As various jurisdictions enact legislation on AI safety, the concept of alignment must be defined and measured across different domains. This paper proposes an experimental framework to assess whether LLMs adhere to ethical and legal standards in the relatively unexplored context of finance. We prompt nine LLMs to impersonate the CEO of a financial institution and test their willingness to misuse customer assets to repay outstanding corporate debt. Beginning with a baseline configuration, we adjust preferences, incentives and constraints, analyzing the impact of each adjustment with logistic regression. Our findings reveal significant heterogeneity in the baseline propensity for unethical behavior of LLMs. Factors such as risk aversion, profit expectations, and regulatory environment consistently influence misalignment in ways predicted by economic theory, although the magnitude of these effects varies across LLMs. This paper highlights both the benefits and limitations of simulation-based, ex post safety testing. While it can inform financial authorities and institutions aiming to ensure LLM safety, there is a clear trade-off between generality and cost.

As LLMs become more widely deployed, there is increasing interest in directly optimizing for feedback from end users (e.g. thumbs up) in addition to feedback from paid annotators. However, training to maximize human feedback creates a perverse incentive structure for the AI to resort to manipulative or deceptive tactics to obtain positive feedback from users who are vulnerable to such strategies. We study this phenomenon by training LLMs with Reinforcement Learning with simulated user feedback in environments of practical LLM usage. In our settings, we find that: 1) Extreme forms of "feedback gaming" such as manipulation and deception are learned reliably; 2) Even if only 2% of users are vulnerable to manipulative strategies, LLMs learn to identify and target them while behaving appropriately with other users, making such behaviors harder to detect; 3) To mitigate this issue, it may seem promising to leverage continued safety training or LLM-as-judges during training to filter problematic outputs. Instead, we found that while such approaches help in some of our settings, they backfire in others, sometimes even leading to subtler manipulative behaviors. We hope our results can serve as a case study which highlights the risks of using gameable feedback sources -- such as user feedback -- as a target for RL.

Due to the high cost of retraining from scratch, researchers attempt to employ machine unlearning to remove specific content from LLMs while preserving the overall performance. In this paper, we discuss several issues in machine unlearning for LLMs and provide our insights on possible approaches. To address the issue of inadequate evaluation of model outputs after unlearning, we introduce three additional metrics to evaluate token diversity, sentence semantics, and factual correctness. We then categorize unlearning methods into untargeted and targeted, and discuss their issues respectively. Specifically, the behavior that untargeted unlearning attempts to approximate is unpredictable and may involve hallucinations, and existing regularization is insufficient for targeted unlearning. To alleviate these issues, we propose using the objective of maximizing entropy (ME) for untargeted unlearning and incorporate answer preservation (AP) loss as regularization for targeted unlearning. Experimental results across three scenarios, i.e., fictitious unlearning, continual unlearning, and real-world unlearning, demonstrate the effectiveness of our approaches. In recent years, large language models (LLMs) have undergone rapid development, demonstrating impressive capabilities across a wide range of applications, from natural language processing to complex problem-solving. These concerns are particularly relevant within legal and regulatory frameworks, such as the Right to be Forgotten (Dang, 2021), which aims to empower individuals to have unauthorized data erased from digital records. Addressing these issues is crucial for ensuring the responsible deployment of LLMs in real-world applications. Due to the high cost of retraining LLMs, researchers have explored machine unlearning techniques, namely LLM unlearning (Cao & Yang, 2015; Bourtoule et al., 2021; Yao et al., 2023). The typical paradigm involves fine-tuning the target LLM on a specified set, known as the forget set, to obtain an unlearned model. As described in (Maini et al., 2024; Jin et al., 2024), the unlearned model should meet two primary goals: 1) it should not reveal any information contained in the forget set, and 2) it should maintain performance on the neighbor set, which has a distribution similar to the forget set but is not the target of unlearning, as well as on other tasks with general knowledge. While the first goal is generally easier to achieve, the main challenge lies in meeting the second goal (Liu et al., 2024b; Maini et al., 2024; Zhang et al., 2024a; Ji et al., 2024; Shi et al., 2024a; Wang et al., 2024c). In this paper, we have a closer look at machine unlearning for LLMs. We note that most prior studies (Maini et al., 2024; Ji et al., 2024; Jia et al., 2024; Jin et al., 2024; Shi et al., 2024a) primarily rely on ROUGE (Lin, 2004) as the sole metric for evaluating the output of unlearned models.

Urban air pollution hotspots pose significant health risks, yet their detection and analysis remain limited by the sparsity of public sensor networks. This paper addresses this challenge by combining predictive modeling and mechanistic approaches to comprehensively monitor pollution hotspots. We enhanced New Delhi's existing sensor network with 28 low-cost sensors, collecting PM2.5 data over 30 months from May 1, 2018, to Nov 1, 2020. Applying established definitions of hotspots to this data, we found the existence of additional 189 hidden hotspots apart from confirming 660 hotspots detected by the public network. Using predictive techniques like Space-Time Kriging, we identified hidden hotspots with 95% precision and 88% recall with 50% sensor failure rate, and with 98% precision and 95% recall with 50% missing sensors. The projected results of our predictive models were further compiled into policy recommendations for public authorities. Additionally, we developed a Gaussian Plume Dispersion Model to understand the mechanistic underpinnings of hotspot formation, incorporating an emissions inventory derived from local sources. Our mechanistic model is able to explain 65% of observed transient hotspots. Our findings underscore the importance of integrating data-driven predictive models with physics-based mechanistic models for scalable and robust air pollution management in resource-constrained settings.

The US Patent and Trademark Office banned the use of generative artificial intelligence for any purpose last year, citing security concerns with the technology as well as the propensity of some tools to exhibit "bias, unpredictability, and malicious behavior," according to an April 2023 internal guidance memo obtained by WIRED through a public records request. Jamie Holcombe, the chief information officer of the USPTO, wrote that the office is "committed to pursuing innovation within our agency" but are still "working to bring these capabilities to the office in a responsible way." Paul Fucito, press secretary for the USPTO, clarified to WIRED that employees can use "state-of-the-art generative AI models" at work--but only inside the agency's internal testing environment. "Innovators from across the USPTO are now using the AI Lab to better understand generative AI's capabilities and limitations and to prototype AI-powered solutions to critical business needs," Fucito wrote in an email. Outside of the testing environment, USPTO staff are barred from relying on AI programs like OpenAI's ChatGPT or Anthropic's Claude for work tasks.

As large language models (LLMs) become integrated to decision-making across various sectors, a key question arises: do they exhibit an emergent "moral mind" - a consistent set of moral principles guiding their ethical judgments - and is this reasoning uniform or diverse across models? To investigate this, we presented about forty different models from the main providers with a large array of structured ethical scenarios, creating one of the largest datasets of its kind. Our rationality tests revealed that at least one model from each provider demonstrated behavior consistent with stable moral principles, effectively acting as approximately optimizing a utility function encoding ethical reasoning. We identified these utility functions and observed a notable clustering of models around neutral ethical stances. To investigate variability, we introduced a novel non-parametric permutation approach, revealing that the most rational models shared 59% to 76% of their ethical reasoning patterns. Despite this shared foundation, differences emerged: roughly half displayed greater moral adaptability, bridging diverse perspectives, while the remainder adhered to more rigid ethical structures.

AI models are increasingly prevalent in high-stakes environments, necessitating thorough assessment of their capabilities and risks. Benchmarks are popular for measuring these attributes and for comparing model performance, tracking progress, and identifying weaknesses in foundation and non-foundation models. They can inform model selection for downstream tasks and influence policy initiatives. However, not all benchmarks are the same: their quality depends on their design and usability. In this paper, we develop an assessment framework considering 46 best practices across an AI benchmark's lifecycle and evaluate 24 AI benchmarks against it. We find that there exist large quality differences and that commonly used benchmarks suffer from significant issues. We further find that most benchmarks do not report statistical significance of their results nor allow for their results to be easily replicated. To support benchmark developers in aligning with best practices, we provide a checklist for minimum quality assurance based on our assessment. We also develop a living repository of benchmark assessments to support benchmark comparability, accessible at betterbench.stanford.edu.

The rapid advancement of artificial intelligence has led to increasingly sophisticated deep learning models, which frequently operate as opaque 'black boxes' with limited transparency in their decision-making processes. This lack of interpretability presents considerable challenges, especially in high-stakes applications where understanding the rationale behind a model's outputs is as essential as the outputs themselves. This study addresses the pressing need for interpretability in AI systems, emphasizing its role in fostering trust, ensuring accountability, and promoting responsible deployment in mission-critical fields. To address the interpretability challenge in deep learning, we introduce DLBacktrace, an innovative technique developed by the AryaXAI team to illuminate model decisions across a wide array of domains, including simple Multi Layer Perceptron (MLPs), Convolutional Neural Networks (CNNs), Large Language Models (LLMs), Computer Vision Models, and more. We provide a comprehensive overview of the DLBacktrace algorithm and present benchmarking results, comparing its performance against established interpretability methods, such as SHAP, LIME, GradCAM, Integrated Gradients, SmoothGrad, and Attention Rollout, using diverse task-based metrics. The proposed DLBacktrace technique is compatible with various model architectures built in PyTorch and TensorFlow, supporting models like Llama 3.2, other NLP architectures such as BERT and LSTMs, computer vision models like ResNet and U-Net, as well as custom deep neural network (DNN) models for tabular data. This flexibility underscores DLBacktrace's adaptability and effectiveness in enhancing model transparency across a broad spectrum of applications. The library is open-sourced and available at https://github.com/AryaXAI/DLBacktrace .

Triple Entry (TE) is an accounting method that utilizes three accounts or 'entries' to record each transaction, rather than the conventional double-entry bookkeeping system. Existing studies have found that TE accounting, with its additional layer of verification and disclosure of inter-organizational relationships, could help improve transparency in complex financial and supply chain transactions such as blockchain. Machine learning (ML) presents a promising avenue to augment the transparency advantages of TE accounting. By automating some of the data collection and analysis needed for TE bookkeeping, ML techniques have the potential to make this more transparent accounting method scalable for large organizations with complex international supply chains, further enhancing the visibility and trustworthiness of financial reporting. By leveraging ML algorithms, anomalies within distributed ledger data can be swiftly identified, flagging potential instances of fraud or errors. Furthermore, by delving into transaction relationships over time, ML can untangle intricate webs of transactions, shedding light on obscured dealings and adding an investigative dimension. This paper aims to demonstrate the interaction between TE and ML and how they can leverage transparency levels.