arXiv subjects

Shu Yang

Publications and source records attributed to Shu Yang.

At least 181 records · Page 10Linked to original sources

arXiv2024

A Hopfieldian View-based Interpretation for Chain-of-Thought Reasoning

Chain-of-Thought (CoT) holds a significant place in augmenting the reasoning performance for large language models (LLMs). While some studies focus on improving CoT accuracy through methods like retrieval enhancement, yet a rigorous explanation for why CoT achieves such success remains unclear. In this paper, we analyze CoT methods under two different settings by asking the following questions: (1) For zero-shot CoT, why does prompting the model with "let's think step by step" significantly impact its outputs? (2) For few-shot CoT, why does providing examples before questioning the model could substantially improve its reasoning ability? To answer these questions, we conduct a top-down explainable analysis from the Hopfieldian view and propose a Read-and-Control approach for controlling the accuracy of CoT. Through extensive experiments on seven datasets for three different tasks, we demonstrate that our framework can decipher the inner workings of CoT, provide reasoning error localization, and control to come up with the correct reasoning path.

cs.CL↗

arXiv2024

A Practical Analysis Procedure on Generalizing Comparative Effectiveness in the Randomized Clinical Trial to the Real-world Trialeligible Population

When evaluating the effectiveness of a drug, a Randomized Controlled Trial (RCT) is often considered the gold standard due to its perfect randomization. While RCT assures strong internal validity, its restricted external validity poses challenges in extending treatment effects to the broader real-world population due to possible heterogeneity in covariates. In this paper, we introduce a procedure to generalize the RCT findings to the real-world trial-eligible population based on the adaption of existing statistical methods. We utilized the augmented inversed probability of sampling weighting (AIPSW) estimator for the estimation and omitted variable bias framework to assess the robustness of the estimate against the assumption violation caused by potentially unmeasured confounders. We analyzed an RCT comparing the effectiveness of lowering hypertension between Songling Xuemaikang Capsule (SXC), a traditional Chinese medicine (TCM), and Losartan as an illustration. The generalization results indicated that although SXC is less effective in lowering blood pressure than Losartan on week 2, week 4, and week 6, there is no statistically significant difference among the trial-eligible population at week 8, and the generalization is robust against potential unmeasured confounders.

stat.AP↗

arXiv2024

Leveraging Logical Rules in Knowledge Editing: A Cherry on the Top

Multi-hop Question Answering (MQA) under knowledge editing (KE) is a key challenge in Large Language Models (LLMs). While best-performing solutions in this domain use a plan and solve paradigm to split a question into sub-questions followed by response generation, we claim that this approach is sub-optimal as it fails for hard to decompose questions, and it does not explicitly cater to correlated knowledge updates resulting as a consequence of knowledge edits. This has a detrimental impact on the overall consistency of the updated knowledge. To address these issues, in this paper, we propose a novel framework named RULE-KE, i.e., RULE based Knowledge Editing, which is a cherry on the top for augmenting the performance of all existing MQA methods under KE. Specifically, RULE-KE leverages rule discovery to discover a set of logical rules. Then, it uses these discovered rules to update knowledge about facts highly correlated with the edit. Experimental evaluation using existing and newly curated datasets (i.e., RKE-EVAL) shows that RULE-KE helps augment both performances of parameter-based and memory-based solutions up to 92% and 112.9%, respectively.

cs.CL↗

arXiv2024

Inference for Optimal Linear Treatment Regimes in Personalized Decision-making

Personalized decision-making, tailored to individual characteristics, is gaining significant attention. The optimal treatment regime aims to provide the best-expected outcome in the entire population, known as the value function. One approach to determine this optimal regime is by maximizing the Augmented Inverse Probability Weighting (AIPW) estimator of the value function. However, the derived treatment regime can be intricate and nonlinear, limiting their use. For clarity and interoperability, we emphasize linear regimes and determine the optimal linear regime by optimizing the AIPW estimator within set constraints. While the AIPW estimator offers a viable path to estimating the optimal regime, current methodologies predominantly focus on its asymptotic distribution, leaving a gap in studying the linear regime itself. However, there are many benefits to understanding the regime, as pinpointing significant covariates can enhance treatment effects and provide future clinical guidance. In this paper, we explore the asymptotic distribution of the estimated linear regime. Our results show that the parameter associated with the linear regime follows a cube-root convergence to a non-normal limiting distribution characterized by the maximizer of a centered Gaussian process with a quadratic drift. When making inferences for the estimated linear regimes with cube-root convergence in practical scenarios, the standard nonparametric bootstrap is invalid. As a solution, we facilitate the Cattaneo et al. (2020) bootstrap technique to provide a consistent distributional approximation for the estimated linear regimes, validated further through simulations and real-world data applications from the eICU Collaborative Research Database.

stat.ME↗

arXiv2024

Causal Customer Churn Analysis with Low-rank Tensor Block Hazard Model

This study introduces an innovative method for analyzing the impact of various interventions on customer churn, using the potential outcomes framework. We present a new causal model, the tensorized latent factor block hazard model, which incorporates tensor completion methods for a principled causal analysis of customer churn. A crucial element of our approach is the formulation of a 1-bit tensor completion for the parameter tensor. This captures hidden customer characteristics and temporal elements from churn records, effectively addressing the binary nature of churn data and its time-monotonic trends. Our model also uniquely categorizes interventions by their similar impacts, enhancing the precision and practicality of implementing customer retention strategies. For computational efficiency, we apply a projected gradient descent algorithm combined with spectral clustering. We lay down the theoretical groundwork for our model, including its non-asymptotic properties. The efficacy and superiority of our model are further validated through comprehensive experiments on both simulated and real-world applications.

stat.ML↗

arXiv2024

A Survey on LLM-Generated Text Detection: Necessity, Methods, and Future Directions

The powerful ability to understand, follow, and generate complex language emerging from large language models (LLMs) makes LLM-generated text flood many areas of our daily lives at an incredible speed and is widely accepted by humans. As LLMs continue to expand, there is an imperative need to develop detectors that can detect LLM-generated text. This is crucial to mitigate potential misuse of LLMs and safeguard realms like artistic expression and social networks from harmful influence of LLM-generated content. The LLM-generated text detection aims to discern if a piece of text was produced by an LLM, which is essentially a binary classification task. The detector techniques have witnessed notable advancements recently, propelled by innovations in watermarking techniques, statistics-based detectors, neural-base detectors, and human-assisted methods. In this survey, we collate recent research breakthroughs in this area and underscore the pressing need to bolster detector research. We also delve into prevalent datasets, elucidating their limitations and developmental requirements. Furthermore, we analyze various LLM-generated text detection paradigms, shedding light on challenges like out-of-distribution problems, potential attacks, real-world data issues and the lack of effective evaluation framework. Conclusively, we highlight interesting directions for future research in LLM-generated text detection to advance the implementation of responsible artificial intelligence (AI). Our aim with this survey is to provide a clear and comprehensive introduction for newcomers while also offering seasoned researchers a valuable update in the field of LLM-generated text detection. The useful resources are publicly available at: https://github.com/NLP2CT/LLM-generated-Text-Detection.

cs.CL↗

arXiv2024

MONAL: Model Autophagy Analysis for Modeling Human-AI Interactions

The increasing significance of large models and their multi-modal variants in societal information processing has ignited debates on social safety and ethics. However, there exists a paucity of comprehensive analysis for: (i) the interactions between human and artificial intelligence systems, and (ii) understanding and addressing the associated limitations. To bridge this gap, we propose Model Autophagy Analysis (MONAL) for large models' self-consumption explanation. MONAL employs two distinct autophagous loops (referred to as ``self-consumption loops'') to elucidate the suppression of human-generated information in the exchange between human and AI systems. Through comprehensive experiments on diverse datasets, we evaluate the capacities of generated models as both creators and disseminators of information. Our key findings reveal (i) A progressive prevalence of model-generated synthetic information over time within training datasets compared to human-generated information; (ii) The discernible tendency of large models, when acting as information transmitters across multiple iterations, to selectively modify or prioritize specific contents; and (iii) The potential for a reduction in the diversity of socially or human-generated information, leading to bottlenecks in the performance enhancement of large models and confining them to local optima.

cs.CL↗

arXiv2024

Dialectical Alignment: Resolving the Tension of 3H and Security Threats of LLMs

With the rise of large language models (LLMs), ensuring they embody the principles of being helpful, honest, and harmless (3H), known as Human Alignment, becomes crucial. While existing alignment methods like RLHF, DPO, etc., effectively fine-tune LLMs to match preferences in the preference dataset, they often lead LLMs to highly receptive human input and external evidence, even when this information is poisoned. This leads to a tendency for LLMs to be Adaptive Chameleons when external evidence conflicts with their parametric memory. This exacerbates the risk of LLM being attacked by external poisoned data, which poses a significant security risk to LLM system applications such as Retrieval-augmented generation (RAG). To address the challenge, we propose a novel framework: Dialectical Alignment (DA), which (1) utilizes AI feedback to identify optimal strategies for LLMs to navigate inter-context conflicts and context-memory conflicts with different external evidence in context window (i.e., different ratios of poisoned factual contexts); (2) constructs the SFT dataset as well as the preference dataset based on the AI feedback and strategies above; (3) uses the above datasets for LLM alignment to defense poisoned context attack while preserving the effectiveness of in-context knowledge editing. Our experiments show that the dialectical alignment model improves poisoned data attack defense by 20 and does not require any additional prompt engineering or prior declaration of ``you may be attacked`` to the LLMs' context window.

cs.CL↗

arXiv2024

MambaMIL: Enhancing Long Sequence Modeling with Sequence Reordering in Computational Pathology

Multiple Instance Learning (MIL) has emerged as a dominant paradigm to extract discriminative feature representations within Whole Slide Images (WSIs) in computational pathology. Despite driving notable progress, existing MIL approaches suffer from limitations in facilitating comprehensive and efficient interactions among instances, as well as challenges related to time-consuming computations and overfitting. In this paper, we incorporate the Selective Scan Space State Sequential Model (Mamba) in Multiple Instance Learning (MIL) for long sequence modeling with linear complexity, termed as MambaMIL. By inheriting the capability of vanilla Mamba, MambaMIL demonstrates the ability to comprehensively understand and perceive long sequences of instances. Furthermore, we propose the Sequence Reordering Mamba (SR-Mamba) aware of the order and distribution of instances, which exploits the inherent valuable information embedded within the long sequences. With the SR-Mamba as the core component, MambaMIL can effectively capture more discriminative features and mitigate the challenges associated with overfitting and high computational overhead. Extensive experiments on two public challenging tasks across nine diverse datasets demonstrate that our proposed framework performs favorably against state-of-the-art MIL methods. The code is released at https://github.com/isyangshu/MambaMIL.

cs.CV↗

arXiv2024

SDiT: Spiking Diffusion Model with Transformer

Spiking neural networks (SNNs) have low power consumption and bio-interpretable characteristics, and are considered to have tremendous potential for energy-efficient computing. However, the exploration of SNNs on image generation tasks remains very limited, and a unified and effective structure for SNN-based generative models has yet to be proposed. In this paper, we explore a novel diffusion model architecture within spiking neural networks. We utilize transformer to replace the commonly used U-net structure in mainstream diffusion models. It can generate higher quality images with relatively lower computational cost and shorter sampling time. It aims to provide an empirical baseline for research of generative models based on SNNs. Experiments on MNIST, Fashion-MNIST, and CIFAR-10 datasets demonstrate that our work is highly competitive compared to existing SNN generative models.

cs.CV↗

arXiv2024

MoRAL: MoE Augmented LoRA for LLMs' Lifelong Learning

Adapting large language models (LLMs) to new domains/tasks and enabling them to be efficient lifelong learners is a pivotal challenge. In this paper, we propose MoRAL, i.e., Mixture-of-Experts augmented Low-Rank Adaptation for Lifelong Learning. MoRAL combines the multi-tasking abilities of MoE with the fine-tuning abilities of LoRA for effective life-long learning of LLMs. In contrast to the conventional approaches that use factual triplets as inputs MoRAL relies on simple question-answer pairs, which is a more practical and effective strategy for robust and efficient learning. Owing to new data settings, we introduce a new evaluation benchmark namely: Life Long Learning of LLM (5L-bench) encompassing a newly curated dataset of question-answer pairs, and a set of evaluation metrics for rigorous evaluation of MoRAL in open-book and closed-book settings. Experimental evaluation shows (i) LLMs learn fast in open-book settings with up to 30.15% improvement in "RA" for Phi-2-2.7B compared to closed-book (for models fine-tuned with MoRAL); (ii) MoRAL shows higher performance improvement for models with a greater number of parameters; (iii) MoRAL is robust to catastrophic forgetting offering better knowledge retention compared to baselines.

cs.CL↗

arXiv2024

Mixed Matrix Completion in Complex Survey Sampling under Heterogeneous Missingness

Modern surveys with large sample sizes and growing mixed-type questionnaires require robust and scalable analysis methods. In this work, we consider recovering a mixed dataframe matrix, obtained by complex survey sampling, with entries following different canonical exponential distributions and subject to heterogeneous missingness. To tackle this challenging task, we propose a two-stage procedure: in the first stage, we model the entry-wise missing mechanism by logistic regression, and in the second stage, we complete the target parameter matrix by maximizing a weighted log-likelihood with a low-rank constraint. We propose a fast and scalable estimation algorithm that achieves sublinear convergence, and the upper bound for the estimation error of the proposed method is rigorously derived. Experimental results support our theoretical claims, and the proposed estimator shows its merits compared to other existing methods. The proposed method is applied to analyze the National Health and Nutrition Examination Survey data.

stat.ME↗

arXiv2024

Propensity score matching for estimating a marginal hazard ratio

Propensity score matching is commonly used to draw causal inference from observational survival data. However, its asymptotic properties have yet to be established, and variance estimation is still open to debate. We derive the statistical properties of the propensity score matching estimator of the marginal causal hazard ratio based on matching with replacement and a fixed number of matches. We also propose a double-resampling technique for variance estimation that takes into account the uncertainty due to propensity score estimation prior to matching.

stat.ME↗

arXiv2024

Continuous-time structural failure time model for intermittent treatment

The intermittent intake of treatment is commonly seen in patients with chronic disease. For example, patients with atrial fibrillation may need to discontinue the oral anticoagulants when they experience a certain surgery and re-initiate the treatment after the surgery. As another example, patients may skip a few days before they refill a treatment as planned. This treatment dispensation information (i.e., the time at which a patient initiates and refills a treatment) is recorded in the electronic healthcare records or claims database, and each patient has a different treatment dispensation. Current methods to estimate the effects of such treatments censor the patients who re-initiate the treatment, which results in information loss or biased estimation. In this work, we present methods to estimate the effect of treatments on failure time outcomes by taking all the treatment dispensation information. The developed methods are based on the continuous-time structural failure time model, where the dependent censoring is tackled by inverse probability of censoring weighting. The estimators are doubly robust and locally efficient.

stat.ME↗

arXiv2024

Second harmonic generation at a time-varying interface

Time-varying metamaterials rely on large and fast changes of the linear permittivity. Beyond the linear terms, however, the effect of a non-perturbative modulation of the medium on harmonic generation and the associated nonlinear susceptibilities remains largely unexplored. In this work, we study second harmonic generation at an optically pumped time-varying interface between air and a 310 nm Indium Tin Oxide film. We observe an enhancement of the modulation contrast at the second harmonic wavelength, up to 93% for a pump intensity of 100 GW/cm$^2$, leading to large frequency broadening and shift. We demonstrate that, in addition to the quadratic dependence on the fundamental field, a significant contribution to the enhancement comes from the temporal modulation of the second order nonlinear susceptibility, whose relative change is double that of the linear term. Moreover, the spectra resulting from single and double-slit time diffraction show significantly enhanced frequency shift, broadening and modulation depth, when compared to the infrared fundamental beam, and could be exploited for optical computing and sensing. Enhanced time-varying effects on the harmonic signal extends the application of materials to the visible range and calls for further theoretical exploration of non-perturbative nonlinear optics.

physics.optics↗

arXiv2023

Multiply robust estimators in longitudinal studies with missing data under control-based imputation

Longitudinal studies are often subject to missing data. The ICH E9(R1) addendum addresses the importance of defining a treatment effect estimand with the consideration of intercurrent events. Jump-to-reference (J2R) is one classically envisioned control-based scenario for the treatment effect evaluation using the hypothetical strategy, where the participants in the treatment group after intercurrent events are assumed to have the same disease progress as those with identical covariates in the control group. We establish new estimators to assess the average treatment effect based on a proposed potential outcomes framework under J2R. Various identification formulas are constructed under the assumptions addressed by J2R, motivating estimators that rely on different parts of the observed data distribution. Moreover, we obtain a novel estimator inspired by the efficient influence function, with multiple robustness in the sense that it achieves $n^{1/2}$-consistency if any pairs of multiple nuisance functions are correctly specified, or if the nuisance functions converge at a rate not slower than $n^{-1/4}$ when using flexible modeling approaches. The finite-sample performance of the proposed estimators is validated in simulation studies and an antidepressant clinical trial.

stat.ME↗

arXiv2023

Positivity-free Policy Learning with Observational Data

Policy learning utilizing observational data is pivotal across various domains, with the objective of learning the optimal treatment assignment policy while adhering to specific constraints such as fairness, budget, and simplicity. This study introduces a novel positivity-free (stochastic) policy learning framework designed to address the challenges posed by the impracticality of the positivity assumption in real-world scenarios. This framework leverages incremental propensity score policies to adjust propensity score values instead of assigning fixed values to treatments. We characterize these incremental propensity score policies and establish identification conditions, employing semiparametric efficiency theory to propose efficient estimators capable of achieving rapid convergence rates, even when integrated with advanced machine learning algorithms. This paper provides a thorough exploration of the theoretical guarantees associated with policy learning and validates the proposed framework's finite-sample performance through comprehensive numerical experiments, ensuring the identification of causal effects from observational data is both robust and reliable.

stat.ME↗

arXiv2023

Real Effect or Bias? Best Practices for Evaluating the Robustness of Real-World Evidence through Quantitative Sensitivity Analysis for Unmeasured Confounding

The assumption of no unmeasured confounders is a critical but unverifiable assumption required for causal inference yet quantitative sensitivity analyses to assess robustness of real-world evidence remains underutilized. The lack of use is likely in part due to complexity of implementation and often specific and restrictive data requirements required for application of each method. With the advent of sensitivity analyses methods that are broadly applicable in that they do not require identification of a specific unmeasured confounder, along with publicly available code for implementation, roadblocks toward broader use are decreasing. To spur greater application, here we present a best practice guidance to address the potential for unmeasured confounding at both the design and analysis stages, including a set of framing questions and an analytic toolbox for researchers. The questions at the design stage guide the research through steps evaluating the potential robustness of the design while encouraging gathering of additional data to reduce uncertainty due to potential confounding. At the analysis stage, the questions guide researchers to quantifying the robustness of the observed result and providing researchers with a clearer indication of the robustness of their conclusions. We demonstrate the application of the guidance using simulated data based on a real-world fibromyalgia study, applying multiple methods from our analytic toolbox for illustration purposes.

stat.ME↗