We propose BlockFound, a customized foundation model for anomaly blockchain transaction detection. Unlike existing methods that rely on rule-based systems or directly apply off-the-shelf large language models, BlockFound introduces a series of customized designs to model the unique data structure of blockchain transactions. First, a blockchain transaction is multi-modal, containing blockchain-specific tokens, texts, and numbers. We design a modularized tokenizer to handle these multi-modal inputs, balancing the information across different modalities. Second, we design a customized mask language learning mechanism for pretraining with RoPE embedding and FlashAttention for handling longer sequences. Extensive evaluations on Ethereum and Solana transactions demonstrate BlockFound's exceptional capability in anomaly detection while maintaining a low false positive rate. Remarkably, BlockFound is the only method that successfully detects anomalous transactions on Solana with high accuracy, whereas all other approaches achieved very low or zero detection recall scores. This work not only provides new foundation models for blockchain but also sets a new benchmark for applying LLMs in blockchain data. With the rapid development of blockchain technology, cryptocurrencies have gained significant attention and are increasingly being used in real-world applications. A lot of Decentralized Finance (DeFi) protocols have emerged, offering a wide range of financial services, such as lending, borrowing, and trading, to users. However, the decentralized nature of these protocols also makes them vulnerable to various security threats, including the presence of malicious attacks such as doublespending attack (Karame et al., 2012), partition attacks (Saad et al., 2019), and front-running attacks (Eskandari et al., 2020). These attacks seriously threaten the asset security of billions of blockchain users.

Fine-tuning is widely used to adapt language models for specific goals, often leveraging real-world data such as patient records, customer-service interactions, or web content in languages not covered in pre-training. These datasets are typically massive, noisy, and often confidential, making their direct inspection challenging. However, understanding them is essential for guiding model deployment and informing decisions about data cleaning or suppressing any harmful behaviors learned during fine-tuning. In this study, we introduce the task of novelty discovery through generation, which aims to identify novel properties of a fine-tuning dataset by generating examples that illustrate these properties. Our approach, Contrastive Generative Exploration (CGE), assumes no direct access to the data but instead relies on a pre-trained model and the same model after fine-tuning. By contrasting the predictions of these two models, CGE can generate examples that highlight novel characteristics of the fine-tuning data. However, this simple approach may produce examples that are too similar to one another, failing to capture the full range of novel phenomena present in the dataset. We address this by introducing an iterative version of CGE, where the previously generated examples are used to update the pre-trained model, and this updated model is then contrasted with the fully fine-tuned model to generate the next example, promoting diversity in the generated outputs. Our experiments demonstrate the effectiveness of CGE in detecting novel content, such as toxic language, as well as new natural and programming languages. Furthermore, we show that CGE remains effective even when models are fine-tuned using differential privacy techniques.

Early identification of patients at risk of cardiovascular diseases (CVD) is crucial for effective preventive care, reducing healthcare burden, and improving patients' quality of life. This study demonstrates the potential of retinal optical coherence tomography (OCT) imaging combined with fundus photographs for identifying future adverse cardiac events. We used data from 977 patients who experienced CVD within a 5-year interval post-image acquisition, alongside 1,877 control participants without CVD, totaling 2,854 subjects. We propose a novel binary classification network based on a Multi-channel Variational Autoencoder (MCVAE), which learns a latent embedding of patients' fundus and OCT images to classify individuals into two groups: those likely to develop CVD in the future and those who are not. Our model, trained on both imaging modalities, achieved promising results (AUROC 0.78 +/- 0.02, accuracy 0.68 +/- 0.002, precision 0.74 +/- 0.02, sensitivity 0.73 +/- 0.02, and specificity 0.68 +/- 0.01), demonstrating its efficacy in identifying patients at risk of future CVD events based on their retinal images. This study highlights the potential of retinal OCT imaging and fundus photographs as cost-effective, non-invasive alternatives for predicting cardiovascular disease risk. The widespread availability of these imaging techniques in optometry practices and hospitals further enhances their potential for large-scale CVD risk screening. Our findings contribute to the development of standardized, accessible methods for early CVD risk identification, potentially improving preventive care strategies and patient outcomes.

Large Vision-Language Models (LVLMs) demonstrate impressive capabilities in generating detailed and coherent responses from visual inputs. However, they are prone to generate hallucinations due to an over-reliance on language priors. To address this issue, we investigate the language priors in LVLMs and make two key observations: (1) Even when predicting the tokens associated with image-related part-of-speech (POS), models increasingly rely on linguistic priors as the token sequences grow, thereby amplifying hallucinations. (2) Methods that directly calibrate LVLM's output distribution to mitigate language priors can lead to a degradation in text quality or even exacerbate hallucinations. Based on these findings, we propose a novel method, Summary-Guided Decoding (SGD). This method naturally encourages the model to focus more on image information by reducing the text context through summaries, while controlling only the image-related POS tokens to maintain text quality. Through experiments, we demonstrate that SGD achieves state-of-the-art performance on object hallucination benchmarks. Furthermore, in terms of the trade-off between precision and recall, SGD achieves Pareto optimality among the existing methods. Lastly, we observe that although existing methods struggle to balance the reduction of object hallucinations with maintaining text quality, SGD demonstrates robustness in handling this challenge.

A central question in human immunology is how a patient's repertoire of T cells impacts disease. Here, we introduce a method to infer the causal effects of T cell receptor (TCR) sequences on patient outcomes using observational TCR repertoire sequencing data and clinical outcomes data. Our approach corrects for unobserved confounders, such as a patient's environment and life history, by using the patient's immature, pre-selection TCR repertoire. The pre-selection repertoire can be estimated from nonproductive TCR data, which is widely available. It is generated by a randomized mutational process, V(D)J recombination, which provides a natural experiment. We show formally how to use the pre-selection repertoire to draw causal inferences, and develop a scalable neural-network estimator for our identification formula. Our method produces an estimate of the effect of interventions that add a specific TCR sequence to patient repertoires. As a demonstration, we use it to analyze the effects of TCRs on COVID-19 severity, uncovering potentially therapeutic TCRs that are (1) observed in patients, (2) bind SARS-CoV-2 antigens in vitro and (3) have strong positive effects on clinical outcomes.

Multimodal Large Language Models (MLLMs) have recently received substantial interest, which shows their emerging potential as general-purpose models for various vision-language tasks. MLLMs involve significant external knowledge within their parameters; however, it is challenging to continually update these models with the latest knowledge, which involves huge computational costs and poor interpretability. Retrieval augmentation techniques have proven to be effective plugins for both LLMs and MLLMs. In this study, we propose multimodal adaptive Retrieval-Augmented Bootstrapping Language-Image Pre-training (RA-BLIP), a novel retrieval-augmented framework for various MLLMs. Considering the redundant information within vision modality, we first leverage the question to instruct the extraction of visual information through interactions with one set of learnable queries, minimizing irrelevant interference during retrieval and generation. Besides, we introduce a pre-trained multimodal adaptive fusion module to achieve question text-to-multimodal retrieval and integration of multimodal knowledge by projecting visual and language modalities into a unified semantic space. Furthermore, we present an Adaptive Selection Knowledge Generation (ASKG) strategy to train the generator to autonomously discern the relevance of retrieved knowledge, which realizes excellent denoising performance. Extensive experiments on open multimodal question-answering datasets demonstrate that RA-BLIP achieves significant performance and surpasses the state-of-the-art retrieval-augmented models.

Large language models (LLMs) have demonstrated impressive performance on several tasks and are increasingly deployed in real-world applications. However, especially in high-stakes settings, it becomes vital to know when the output of an LLM may be unreliable. Depending on whether an answer is trustworthy, a system can then choose to route the question to another expert, or otherwise fall back on a safe default behavior. In this work, we study the extent to which LLMs can reliably indicate confidence in their answers, and how this notion of confidence can translate into downstream accuracy gains. We propose Self-REF, a lightweight training strategy to teach LLMs to express confidence in whether their answers are correct in a reliable manner. Self-REF introduces confidence tokens into the LLM, from which a confidence score can be extracted. Compared to conventional approaches such as verbalizing confidence and examining token probabilities, we demonstrate empirically that confidence tokens show significant improvements in downstream routing and rejection learning tasks.

As relevant examples such as the future criminal detection software [1] show, fairness of AI-based and social domain affecting decision support tools constitutes an important area of research. In this contribution, we investigate the applications of AI to socioeconomically relevant infrastructures such as those of water distribution networks (WDNs), where fairness issues have yet to gain a foothold. To establish the notion of fairness in this domain, we propose an appropriate definition of protected groups and group fairness in WDNs as an extension of existing definitions. We demonstrate that typical methods for the detection of leakages in WDNs are unfair in this sense. Further, we thus propose a remedy to increase the fairness which can be applied even to non-differentiable ensemble classification methods as used in this context.

Large Language Models (LLMs) are increasingly used in a variety of important applications, yet their safety and reliability remain as major concerns. Various adversarial and jailbreak attacks have been proposed to bypass the safety alignment and cause the model to produce harmful responses. We introduce Self-supervised Prompt INjection (SPIN) which can detect and reverse these various attacks on LLMs. As our self-supervised prompt defense is done at inference-time, it is also compatible with existing alignment and adds an additional layer of safety for defense. Our benchmarks demonstrate that our system can reduce the attack success rate by up to 87.9%, while maintaining the performance on benign user requests. In addition, we discuss the situation of an adaptive attacker and show that our method is still resilient against attackers who are aware of our defense.

Statistical watermarking techniques are well-established for sequentially decoded language models (LMs). However, these techniques cannot be directly applied to order-agnostic LMs, as the tokens in order-agnostic LMs are not generated sequentially. In this work, we introduce Pattern-mark, a pattern-based watermarking framework specifically designed for order-agnostic LMs. We develop a Markov-chain-based watermark generator that produces watermark key sequences with high-frequency key patterns. Correspondingly, we propose a statistical pattern-based detection algorithm that recovers the key sequence during detection and conducts statistical tests based on the count of high-frequency patterns. Our extensive evaluations on order-agnostic LMs, such as ProteinMPNN and CMLM, demonstrate Pattern-mark's enhanced detection efficiency, generation quality, and robustness, positioning it as a superior watermarking technique for order-agnostic LMs.