arXiv subjects

Bin Yu

Publications and source records attributed to Bin Yu.

At least 91 records · Page 5Linked to original sources

arXiv2024

Veridical Data Science for Medical Foundation Models

The advent of foundation models (FMs) such as large language models (LLMs) has led to a cultural shift in data science, both in medicine and beyond. This shift involves moving away from specialized predictive models trained for specific, well-defined domain questions to generalist FMs pre-trained on vast amounts of unstructured data, which can then be adapted to various clinical tasks and questions. As a result, the standard data science workflow in medicine has been fundamentally altered; the foundation model lifecycle (FMLC) now includes distinct upstream and downstream processes, in which computational resources, model and data access, and decision-making power are distributed among multiple stakeholders. At their core, FMs are fundamentally statistical models, and this new workflow challenges the principles of Veridical Data Science (VDS), hindering the rigorous statistical analysis expected in transparent and scientifically reproducible data science practices. We critically examine the medical FMLC in light of the core principles of VDS: predictability, computability, and stability (PCS), and explain how it deviates from the standard data science workflow. Finally, we propose recommendations for a reimagined medical FMLC that expands and refines the PCS principles for VDS including considering the computational and accessibility constraints inherent to FMs.

cs.LG↗

arXiv2024

Minimum-Norm Interpolation Under Covariate Shift

Transfer learning is a critical part of real-world machine learning deployments and has been extensively studied in experimental works with overparameterized neural networks. However, even in the simplest setting of linear regression a notable gap still exists in the theoretical understanding of transfer learning. In-distribution research on high-dimensional linear regression has led to the identification of a phenomenon known as \textit{benign overfitting}, in which linear interpolators overfit to noisy training labels and yet still generalize well. This behavior occurs under specific conditions on the source covariance matrix and input data dimension. Therefore, it is natural to wonder how such high-dimensional linear models behave under transfer learning. We prove the first non-asymptotic excess risk bounds for benignly-overfit linear interpolators in the transfer learning setting. From our analysis, we propose a taxonomy of \textit{beneficial} and \textit{malignant} covariate shifts based on the degree of overparameterization. We follow our analysis with empirical studies that show these beneficial and malignant covariate shifts for linear interpolators on real image data, and for fully-connected neural networks in settings where the input data dimension is larger than the training sample size.

cs.LG↗

arXiv2024

The Minimum Information about CLinical Artificial Intelligence Checklist for Generative Modeling Research (MI-CLAIM-GEN)

Recent advances in generative models, including large language models (LLMs), vision language models (VLMs), and diffusion models, have accelerated the field of natural language and image processing in medicine and marked a significant paradigm shift in how biomedical models can be developed and deployed. While these models are highly adaptable to new tasks, scaling and evaluating their usage presents new challenges not addressed in previous frameworks. In particular, the ability of these models to produce useful outputs with little to no specialized training data ("zero-" or "few-shot" approaches), as well as the open-ended nature of their outputs, necessitate the development of new guidelines for robust reporting of clinical generative model research. In response to gaps in standards and best practices for the development of clinical AI tools identified by US Executive Order 141103 and several emerging national networks for clinical AI evaluation, we begin to formalize some of these guidelines by building on the original MI-CLAIM checklist. The new checklist, MI-CLAIM-GEN (Table 1), aims to address differences in training, evaluation, interpretability, and reproducibility of new generative models compared to non-generative ("predictive") AI models. This MI-CLAIM-GEN checklist also seeks to clarify cohort selection reporting with unstructured clinical data and adds additional items on alignment with ethical standards for clinical AI research.

cs.CL↗

arXiv2024

Localizing axial dense emitters based onsingle-helix point spread function andcompressed sensing

Among the approaches in three-dimensional (3D) single molecule localization microscopy, there are several point spread function (PSF) engineering approaches, in which depth information of molecules is encoded in 2D images. Usually,the molecules are excited sparsely in each raw image. The consequence is that the temporal resolution has to be sacrificed. In order to improve temporal resolution and ensure localization accuracy, we propose a method, SH-CS, based on light needle excitation, detection system with single helix-point spread function (SH-PSF), and compressed sensing (CS). Although the SH-CS method still has a limitation about the molecule density, it is suited for relatively dense molecules. For each light needle scanning position, an SH image of excited molecules is processed with CS algorithm to decode their axial information. Simulations demonstrated, for random distributed 1 ~ 15 molecules in depth range of 4 μm, the axial localization accuracy is 12.1 nm ~ 73.5 nm. The feasibility of this method is validated by experimental data.

physics.optics↗

arXiv2024

LoRA+: Efficient Low Rank Adaptation of Large Models

In this paper, we show that Low Rank Adaptation (LoRA) as originally introduced in Hu et al. (2021) leads to suboptimal finetuning of models with large width (embedding dimension). This is due to the fact that adapter matrices A and B in LoRA are updated with the same learning rate. Using scaling arguments for large width networks, we demonstrate that using the same learning rate for A and B does not allow efficient feature learning. We then show that this suboptimality of LoRA can be corrected simply by setting different learning rates for the LoRA adapter matrices A and B with a well-chosen ratio. We call this proposed algorithm LoRA$+$. In our extensive experiments, LoRA$+$ improves performance (1-2 $\%$ improvements) and finetuning speed (up to $\sim$ 2X SpeedUp), at the same computational cost as LoRA.

cs.LG↗

arXiv2024

The indexed links of Non-singular Morse-Smale flows on graph manifolds

We classify the indexed links corresponding to the union of the closed orbits of non-singular Morse-Smale flows on most graph manifolds. We find that each of this kind of indexed links can be obtained by applying a finite steps of operations on a special indexed link, which consists of all of the singular Seifert fibers and some regular Seifert fibers with some precisely described conditions.

math.DS↗

arXiv2024

The Impact of Initialization on LoRA Finetuning Dynamics

In this paper, we study the role of initialization in Low Rank Adaptation (LoRA) as originally introduced in Hu et al. (2021). Essentially, to start from the pretrained model as initialization for finetuning, one can either initialize B to zero and A to random (default initialization in PEFT package), or vice-versa. In both cases, the product BA is equal to zero at initialization, which makes finetuning starts from the pretrained model. These two initialization schemes are seemingly similar. They should in-principle yield the same performance and share the same optimal learning rate. We demonstrate that this is an incorrect intuition and that the first scheme (initializing B to zero and A to random) on average yields better performance compared to the other scheme. Our theoretical analysis shows that the reason behind this might be that the first initialization allows the use of larger learning rates (without causing output instability) compared to the second initialization, resulting in more efficient learning of the first scheme. We validate our results with extensive experiments on LLMs.

cs.LG↗

arXiv2024

Large Stepsize Gradient Descent for Logistic Loss: Non-Monotonicity of the Loss Improves Optimization Efficiency

We consider gradient descent (GD) with a constant stepsize applied to logistic regression with linearly separable data, where the constant stepsize $η$ is so large that the loss initially oscillates. We show that GD exits this initial oscillatory phase rapidly -- in $\mathcal{O}(η)$ steps -- and subsequently achieves an $\tilde{\mathcal{O}}(1 / (ηt) )$ convergence rate after $t$ additional steps. Our results imply that, given a budget of $T$ steps, GD can achieve an accelerated loss of $\tilde{\mathcal{O}}(1/T^2)$ with an aggressive stepsize $η:= Θ( T)$, without any use of momentum or variable stepsize schedulers. Our proof technique is versatile and also handles general classification loss functions (where exponential tails are needed for the $\tilde{\mathcal{O}}(1/T^2)$ acceleration), nonlinear predictors in the neural tangent kernel regime, and online stochastic gradient descent (SGD) with a large stepsize, under suitable separability conditions.

cs.LG↗

arXiv2024

Improving Prototypical Visual Explanations with Reward Reweighing, Reselection, and Retraining

In recent years, work has gone into developing deep interpretable methods for image classification that clearly attributes a model's output to specific features of the data. One such of these methods is the Prototypical Part Network (ProtoPNet), which attempts to classify images based on meaningful parts of the input. While this architecture is able to produce visually interpretable classifications, it often learns to classify based on parts of the image that are not semantically meaningful. To address this problem, we propose the Reward Reweighing, Reselecting, and Retraining (R3) post-processing framework, which performs three additional corrective updates to a pretrained ProtoPNet in an offline and efficient manner. The first two steps involve learning a reward model based on collected human feedback and then aligning the prototypes with human preferences. The final step is retraining, which realigns the base features and the classifier layer of the original model with the updated prototypes. We find that our R3 framework consistently improves both the interpretability and the predictive accuracy of ProtoPNet and its variants.

cs.LG↗

arXiv2024

ED-Copilot: Reduce Emergency Department Wait Time with Language Model Diagnostic Assistance

In the emergency department (ED), patients undergo triage and multiple laboratory tests before diagnosis. This time-consuming process causes ED crowding which impacts patient mortality, medical errors, staff burnout, etc. This work proposes (time) cost-effective diagnostic assistance that leverages artificial intelligence systems to help ED clinicians make efficient and accurate diagnoses. In collaboration with ED clinicians, we use public patient data to curate MIMIC-ED-Assist, a benchmark for AI systems to suggest laboratory tests that minimize wait time while accurately predicting critical outcomes such as death. With MIMIC-ED-Assist, we develop ED-Copilot which sequentially suggests patient-specific laboratory tests and makes diagnostic predictions. ED-Copilot employs a pre-trained bio-medical language model to encode patient information and uses reinforcement learning to minimize ED wait time and maximize prediction accuracy. On MIMIC-ED-Assist, ED-Copilot improves prediction accuracy over baselines while halving average wait time from four hours to two hours. ED-Copilot can also effectively personalize treatment recommendations based on patient severity, further highlighting its potential as a diagnostic assistant. Since MIMIC-ED-Assist is a retrospective benchmark, ED-Copilot is restricted to recommend only observed tests. We show ED-Copilot achieves competitive performance without this restriction as the maximum allowed time increases. Our code is available at https://github.com/cxcscmu/ED-Copilot.

cs.CL↗

arXiv2024

Self orbit equivalences on the Anosov flows on Dehn surgeries on the figure-eight knot

Let $M_k$ ($k\in \mathbb Z$ and $|k|>4$) be the $3$-manifold obtained by doing $k$ Dehn surgery on the figure-eight knot, and $X_t$ be the canonical Anosov flow on $M_k$ that is constructed by Goodman. The main result of this article is that if $|k|\gg 0$, then $Mod (M_k) \cong \mathbb {Z}_2 \oplus \mathbb {Z}_2$ and every element of the mapping class group of $M_k$ can be represented by a self orbit equivalence of $X_t$.

math.DS↗

arXiv2024

Towards Practical Requirement Analysis and Verification: A Case Study on Software IP Components in Aerospace Embedded Systems

IP-based software design is a crucial research field that aims to improve efficiency and reliability by reusing complex software components known as intellectual property (IP) components. To ensure the reusability of these components, particularly in security-sensitive software systems, it is necessary to analyze the requirements and perform formal verification for each IP component. However, converting the requirements of IP components from natural language descriptions to temporal logic and subsequently conducting formal verification demands domain expertise and non-trivial manpower. This paper presents a case study on software IP components derived from aerospace embedded systems, with the objective of automating the requirement analysis and verification process. The study begins by employing Large Language Models to convert unstructured natural language into formal specifications. Subsequently, three distinct verification techniques are employed to ascertain whether the source code meets the extracted temporal logic properties. By doing so, five real-world IP components from the China Academy of Space Technology (CAST) have been successfully verified.

cs.SE↗

arXiv2024

A Signature Based Approach Towards Global Channel Charting with Ultra Low Complexity

Channel charting, an unsupervised learning method that learns a low-dimensional representation from channel information to preserve geometrical property of physical space of user equipments (UEs), has drawn many attentions from both academic and industrial communities, because it can facilitate many downstream tasks, such as indoor localization, UE handover, beam management, and so on. However, many previous works mainly focus on charting that only preserves local geometry and use raw channel information to learn the chart, which do not consider the global geometry and are often computationally intensive and very time-consuming. Therefore, in this paper, a novel signature based approach for global channel charting with ultra low complexity is proposed. By using an iterated-integral based method called signature transform, a compact feature map and a novel distance metric are proposed, which enable channel charting with ultra low complexity and preserving both local and global geometry. We demonstrate the efficacy of our method using synthetic and open-source real-field datasets.

cs.IT↗

arXiv2024

Localizing axial dense emitters based on single-helix point spread function and deep learning

Stimulated Emission Depletion Microscopy (STED) can achieve a spatial resolution as high as several nanometers. As a point scanning imaging method, it requires 3D scanning to complete the imaging of 3D samples. The time-consuming 3D scanning can be compressed into a 2D one in the non-diffracting Bessel-Bessel STED (BB-STED) where samples are effectively excited by an optical needle. However, the image is just the 2D projection, i.e., there is no real axial resolution. Therefore, we propose a method to encode axial information to axially dense emitters by using a detection optical path with single-helix point spread function (SH-PSF), and then predicted the depths of the emitters by means of deep learning. Simulation demonstrated that, for a density 1~ 20 emitters in a depth range of 4 nm, an axial precision of ~35 nm can be achieved. Our method also works for experimental data, and an axial precision of ~63 nm can be achieved.

physics.optics↗

arXiv2024

Diagnosing Transformers: Illuminating Feature Spaces for Clinical Decision-Making

Pre-trained transformers are often fine-tuned to aid clinical decision-making using limited clinical notes. Model interpretability is crucial, especially in high-stakes domains like medicine, to establish trust and ensure safety, which requires human engagement. We introduce SUFO, a systematic framework that enhances interpretability of fine-tuned transformer feature spaces. SUFO utilizes a range of analytic and visualization techniques, including Supervised probing, Unsupervised similarity analysis, Feature dynamics, and Outlier analysis to address key questions about model trust and interpretability. We conduct a case study investigating the impact of pre-training data where we focus on real-world pathology classification tasks, and validate our findings on MedNLI. We evaluate five 110M-sized pre-trained transformer models, categorized into general-domain (BERT, TNLR), mixed-domain (BioBERT, Clinical BioBERT), and domain-specific (PubMedBERT) groups. Our SUFO analyses reveal that: (1) while PubMedBERT, the domain-specific model, contains valuable information for fine-tuning, it can overfit to minority classes when class imbalances exist. In contrast, mixed-domain models exhibit greater resistance to overfitting, suggesting potential improvements in domain-specific model robustness; (2) in-domain pre-training accelerates feature disambiguation during fine-tuning; and (3) feature spaces undergo significant sparsification during this process, enabling clinicians to identify common outlier modes among fine-tuned models as demonstrated in this paper. These findings showcase the utility of SUFO in enhancing trust and safety when using transformers in medicine, and we believe SUFO can aid practitioners in evaluating fine-tuned language models for other applications in medicine and in more critical domains.

cs.CL↗

arXiv2024

Towards Consistent Natural-Language Explanations via Explanation-Consistency Finetuning

Large language models (LLMs) often generate convincing, fluent explanations. However, different from humans, they often generate inconsistent explanations on different inputs. For example, an LLM may generate the explanation "all birds can fly" when answering the question "Can sparrows fly?" but meanwhile answer "no" to the related question "Can penguins fly?". Explanations should be consistent across related examples so that they allow a human to simulate the LLM's decision process on multiple examples. We propose explanation-consistency finetuning (EC-finetuning), a method that adapts LLMs to generate more consistent natural-language explanations on related examples. EC-finetuning involves finetuning LLMs on synthetic data that is carefully constructed to contain consistent explanations. Across a variety of question-answering datasets in various domains, EC-finetuning yields a 10.0% relative explanation consistency improvement on four finetuning datasets, and generalizes to seven out-of-distribution datasets not seen during finetuning (+4.5% relative). Code is available at https://github.com/yandachen/explanation-consistency-finetuning .

cs.CL↗

arXiv2023

Explaining black box text modules in natural language with language models

Large language models (LLMs) have demonstrated remarkable prediction performance for a growing array of tasks. However, their rapid proliferation and increasing opaqueness have created a growing need for interpretability. Here, we ask whether we can automatically obtain natural language explanations for black box text modules. A "text module" is any function that maps text to a scalar continuous value, such as a submodule within an LLM or a fitted model of a brain region. "Black box" indicates that we only have access to the module's inputs/outputs. We introduce Summarize and Score (SASC), a method that takes in a text module and returns a natural language explanation of the module's selectivity along with a score for how reliable the explanation is. We study SASC in 3 contexts. First, we evaluate SASC on synthetic modules and find that it often recovers ground truth explanations. Second, we use SASC to explain modules found within a pre-trained BERT model, enabling inspection of the model's internals. Finally, we show that SASC can generate explanations for the response of individual fMRI voxels to language stimuli, with potential applications to fine-grained brain mapping. All code for using SASC and reproducing results is made available on Github.

cs.AI↗

arXiv2023

Using Experience Classification for Training Non-Markovian Tasks

Unlike the standard Reinforcement Learning (RL) model, many real-world tasks are non-Markovian, whose rewards are predicated on state history rather than solely on the current state. Solving a non-Markovian task, frequently applied in practical applications such as autonomous driving, financial trading, and medical diagnosis, can be quite challenging. We propose a novel RL approach to achieve non-Markovian rewards expressed in temporal logic LTL$_f$ (Linear Temporal Logic over Finite Traces). To this end, an encoding of linear complexity from LTL$_f$ into MDPs (Markov Decision Processes) is introduced to take advantage of advanced RL algorithms. Then, a prioritized experience replay technique based on the automata structure (semantics equivalent to LTL$_f$ specification) is utilized to improve the training process. We empirically evaluate several benchmark problems augmented with non-Markovian tasks to demonstrate the feasibility and effectiveness of our approach.

cs.LG↗