Proceedings of machine learning research最新文献_第4页

RAG without Forgetting: Continual Query-Infused Key Memory. 没有遗忘的RAG：持续查询注入的关键记忆。

Proceedings of machine learning research Pub Date : 2026-01-01

Yuntong Hu, Sha Li, Naren Ramakrishnan, Liang Zhao

{"title":"RAG without Forgetting: Continual Query-Infused Key Memory.","authors":"Yuntong Hu, Sha Li, Naren Ramakrishnan, Liang Zhao","doi":"","DOIUrl":"","url":null,"abstract":"Retrieval-augmented generation (RAG) systems commonly improve robustness via query-time adaptations such as query expansion and iterative retrieval. While effective, these approaches are inherently stateless: adaptations are recomputed for each query and discarded thereafter, precluding cumulative learning and repeatedly incurring inference-time cost. Index-side approaches like key expansion introduce persistence but rely on offline preprocessing or heuristic updates that are weakly aligned with downstream task utility, leading to semantic drift and noise accumulation. We propose Evolving Retrieval Memory (ERM), a training-free framework that transforms transient query-time gains into persistent retrieval improvements. ERM updates the retrieval index through correctness-gated feedback, selectively attributes atomic expansion signals to the document keys they benefit, and progressively evolves keys via stable, norm-bounded updates. We show that query and key expansion are theoretically equivalent under standard similarity functions and prove convergence of ERM's selective updates, amortizing optimal query expansion into a stable index with zero inference-time overhead. Experiments on BEIR and BRIGHT across 13 domains demonstrate consistent gains in retrieval and generation, particularly on reasoning-intensive tasks, at native retrieval speed.","PeriodicalId":74504,"journal":{"name":"Proceedings of machine learning research","volume":"306 ","pages":"45247-45271"},"PeriodicalIF":0.0,"publicationDate":"2026-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13626415/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149119994","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Full-Batch Gradient Descent Outperforms One-Pass SGD: Sample Complexity Separation in Single-Index Learning. 在单指标学习中，全批梯度下降优于单次SGD：样本复杂度分离。

Proceedings of machine learning research Pub Date : 2026-01-01

Filip Kovačević, Hong Chang Ji, Denny Wu, Mahdi Soltanolkotabi, Marco Mondelli

{"title":"Full-Batch Gradient Descent Outperforms One-Pass SGD: Sample Complexity Separation in Single-Index Learning.","authors":"Filip Kovačević, Hong Chang Ji, Denny Wu, Mahdi Soltanolkotabi, Marco Mondelli","doi":"","DOIUrl":"","url":null,"abstract":"It is folklore that reusing training data more than once can improve the statistical efficiency of gradient-based learning. However, beyond linear regression, the theoretical advantage of full-batch gradient descent (GD, which always reuses all the data) over one-pass stochastic gradient descent (online SGD, which uses each data point only once) remains unclear. In this work, we consider learning a <math><mi>d</mi></math> -dimensional single-index model with a quadratic activation, for which it is known that one-pass SGD requires <math><mi>n</mi> <mo>≳</mo> <mi>d</mi> <mspace></mspace> <mtext>log</mtext> <mspace></mspace> <mi>d</mi></math> samples to achieve weak recovery. We first show that this <math><mtext>log</mtext> <mspace></mspace> <mi>d</mi></math> factor in the sample complexity persists for full-batch spherical GD on the correlation loss; however, by simply truncating the activation, full-batch GD exhibits a favorable optimization landscape at <math><mi>n</mi> <mo>≃</mo> <mi>d</mi></math> samples, thereby outperforming one-pass SGD (with the same activation) in statistical efficiency. We complement this result with a trajectory analysis of full-batch GD on the squared loss from small initialization, showing that <math><mi>n</mi> <mo>≳</mo> <mi>d</mi></math> samples and <math><mi>T</mi> <mo>≳</mo> <mtext>log</mtext> <mspace></mspace> <mi>d</mi></math> gradient steps suffice to achieve strong (exact) recovery.","PeriodicalId":74504,"journal":{"name":"Proceedings of machine learning research","volume":"306 ","pages":"60661-60725"},"PeriodicalIF":0.0,"publicationDate":"2026-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13626391/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149119959","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Comparing Computational Pathology Foundation Models using Representational Similarity Analysis. 比较计算病理学基础模型使用代表性相似分析。

Proceedings of machine learning research Pub Date : 2026-01-01

Vaibhav Mishra, William Lotter

{"title":"Comparing Computational Pathology Foundation Models using Representational Similarity Analysis.","authors":"Vaibhav Mishra, William Lotter","doi":"","DOIUrl":"","url":null,"abstract":"Foundation models are increasingly developed in computational pathology (CPath) given their promise in facilitating many downstream tasks. While recent studies have evaluated task performance across models, less is known about the structure and variability of their learned representations. Here, we systematically analyze the representational spaces of six CPath foundation models using techniques popularized in computational neuroscience. The models analyzed span vision-language contrastive learning (CONCH, PLIP, KEEP) and self-distillation (UNI (v2), Virchow (v2), Prov-GigaPath) approaches. Through representational similarity analysis using H&E image patches from TCGA, we find that UNI2 and Virchow2 have the most distinct representational structures, whereas Prov-Gigapath has the highest average similarity across models. Having the same training paradigm (vision-only vs. vision-language) did not guarantee higher representational similarity. The representations of all models showed a high slide-dependence, but relatively low disease-dependence. Stain normalization decreased slide-dependence for all models by a range of 5.5% (CONCH) to 20.5% (PLIP). In terms of intrinsic dimensionality, vision-language models demonstrated relatively compact representations, compared to the more distributed representations of vision-only models. These findings highlight opportunities to improve robustness to slide-specific features, inform model ensembling strategies, and provide insights into how training paradigms shape model representations. Our framework is extendable across medical imaging domains, where probing the internal representations of foundation models can support their effective development and deployment.","PeriodicalId":74504,"journal":{"name":"Proceedings of machine learning research","volume":"297 ","pages":"1007-1022"},"PeriodicalIF":0.0,"publicationDate":"2026-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13345687/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148414246","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Context-Aware Filtering of Unstructured Radiology Reports by Anatomical Region. 基于解剖区域的非结构化放射学报告上下文感知过滤。

Proceedings of machine learning research Pub Date : 2026-01-01

Zakk Heile, Pranav Manjunath, Brian Lerner, Samuel Berchuck, Monica Agrawal, Timothy W Dunn

{"title":"Context-Aware Filtering of Unstructured Radiology Reports by Anatomical Region.","authors":"Zakk Heile, Pranav Manjunath, Brian Lerner, Samuel Berchuck, Monica Agrawal, Timothy W Dunn","doi":"","DOIUrl":"","url":null,"abstract":"Radiology reports contain essential clinical information but often remain in unstructured, free-text formats. Notably, multiple imaging examinations performed simultaneously (such as CT head, facial bones, and cervical spine in trauma cases) may be bundled into a single report that consolidates findings from all studies into one free-text document, written jointly. Because individual sentences may reference ambiguous or overlapping anatomy (e.g., \"there is a fracture\"), sentence-level anatomic classification-filtering a report to retain only findings relevant to a specific anatomical region-is essential for downstream tasks such as structured label extraction and for creating clean, bijective training data for radiology report generation models. While formatting differs across reports, the clinical language remains precise. Using that fact, we develop context-aware classical models with feature engineering that surpass trained neural networks and pre-trained language models. We show that the learned model weights generalize effectively to MIMIC-IV radiology reports and that our approach achieves near-optimal performance with only a small amount of labeled training data. Together, these results make our approach practical and reproducible for new settings.","PeriodicalId":74504,"journal":{"name":"Proceedings of machine learning research","volume":"297 ","pages":"864-885"},"PeriodicalIF":0.0,"publicationDate":"2026-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13364232/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148439358","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Regularizing Extrapolation in Causal Inference. 因果推理中的正则化外推。

Proceedings of machine learning research Pub Date : 2026-01-01

David Arbour, Harsh Parikh, Bijan Niknam, Elizabeth Stuart, Kara Rudolph, Avi Feller

{"title":"Regularizing Extrapolation in Causal Inference.","authors":"David Arbour, Harsh Parikh, Bijan Niknam, Elizabeth Stuart, Kara Rudolph, Avi Feller","doi":"","DOIUrl":"","url":null,"abstract":"Many causal inference and machine learning estimators are linear smoothers, where the prediction is a weighted average of training outcomes. Whether weights are constrained to be non-negative creates a key tradeoff: non-negative weights (e.g., inverse propensity weighting, random forests) limit extrapolation but can worsen covariate imbalance, while unconstrained weights (e.g., OLS, kernel ridge regression) improve balance but increase dependence on parametric assumptions. We propose a unified framework that directly penalizes extrapolation via a soft constraint on negative weights, replacing the standard hard non-negativity restriction. We derive a worst-case error bound and introduce a novel \"bias-bias-variance\" tradeoff among distributional imbalance, model misspecification, and estimator variance; this tradeoff is especially pronounced in high dimensions with poor positivity. We develop a convex optimization procedure that regularizes this bound and outline how to use the extrapolation penalty as a sensitivity analysis for parametric assumptions. We demonstrate our approach on synthetic data and a real-world application generalizing randomized trial estimates to a target population.","PeriodicalId":74504,"journal":{"name":"Proceedings of machine learning research","volume":"300 ","pages":"784-792"},"PeriodicalIF":0.0,"publicationDate":"2026-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13580717/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148992966","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Smooth Multi-Policy Causal Effect Estimation in Longitudinal Settings. 纵向设置下平滑多策略因果效应估计。

Proceedings of machine learning research Pub Date : 2026-01-01

Wenxin Chen, Weishen Pan, Kyra Gan, Fei Wang

{"title":"Smooth Multi-Policy Causal Effect Estimation in Longitudinal Settings.","authors":"Wenxin Chen, Weishen Pan, Kyra Gan, Fei Wang","doi":"","DOIUrl":"","url":null,"abstract":"Comparative evaluation of multiple dynamic treatment policies is essential for healthcare and policy decisions, yet conventional longitudinal causal inference methods estimate each in isolation, preventing information sharing across counterfactuals. We demonstrate that this separate estimation paradigm induces a structurally uncontrolled second-order bias, inflating finite-sample variance even after standard debiasing with longitudinal targeted maximum likelihood estimation (LTMLE). To address this, we propose a policy-aware reparameterization of Iterative Conditional Expectation (ICE) Q-functions that enables joint estimation through shared representations. We implement this approach in the Policy-Encoded Q Network (PEQ-Net), an architecture centered on a shared policy encoder. The encoder is trained using kernel mean embeddings, ensuring that the learned representation space reflects population-level policy dissimilarities. After applying an LTMLE correction step, we prove this design imposes a structural constraint on the second-order remainder, thereby stabilizing finite-sample variance. Experiments on semi-synthetic datasets demonstrate that PEQ-Net consistently outperforms existing ICE-based methods, achieving substantial reductions in root-mean-square error, particularly when evaluating closely related policies.","PeriodicalId":74504,"journal":{"name":"Proceedings of machine learning research","volume":"306 ","pages":"16272-16294"},"PeriodicalIF":0.0,"publicationDate":"2026-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13655700/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149187008","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Asymmetric Prompt Weighting for Reinforcement Learning with Verifiable Rewards. 基于可验证奖励的非对称提示加权强化学习。

Proceedings of machine learning research Pub Date : 2026-01-01

Reinhard Heckel, Mahdi Soltanolkotabi, Christos Thramboulidis

{"title":"Asymmetric Prompt Weighting for Reinforcement Learning with Verifiable Rewards.","authors":"Reinhard Heckel, Mahdi Soltanolkotabi, Christos Thramboulidis","doi":"","DOIUrl":"","url":null,"abstract":"Reinforcement learning with verifiable rewards has driven recent advances in LLM post-training, in particular for reasoning. Policy optimization algorithms generate a number of responses for a given prompt and then effectively weight the corresponding gradients depending on the rewards. The most popular algorithms including GRPO, DAPO, and RLOO focus on ambiguous prompts, i.e., prompts with intermediate success probability, while downgrading gradients with very easy and very hard prompts. In this paper, we consider asymmetric prompt weightings that assign higher weights to prompts with low, or even zero, empirical success probability. We find that asymmetric weighting particularly benefits from-scratch RL (as in R1-Zero), where training traverses a wide accuracy range, and less so in post-SFT RL where the model already starts at high accuracy. We also provide theory that characterizes prompt weights which minimize the time needed to raise success probability from an initial level to a target accuracy under a fixed update budget. In low-success regimes, where informative responses are rare and response cost dominates, these optimal weights become asymmetric, upweighting low success probabilities and thereby accelerating effective-time convergence.","PeriodicalId":74504,"journal":{"name":"Proceedings of machine learning research","volume":"306 ","pages":"42459-42480"},"PeriodicalIF":0.0,"publicationDate":"2026-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13626420/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149119956","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

EventFlow: Forecasting Temporal Point Processes with Flow Matching. EventFlow：用流匹配预测时间点过程。

Proceedings of machine learning research Pub Date : 2026-01-01

Gavin Kerrigan, Kai Nelson, Padhraic Smyth

引用次数: 0

From Zero-Shot to Bedside: A Practical Playbook for Adapting Open-Source Large Language Models to Clinical Symptom Extraction. 从零射击到床边：一个实用的剧本适应开源大语言模型临床症状提取。

Proceedings of machine learning research Pub Date : 2026-01-01

Li-Ching Chen, Travis Zack, Divneet Mandair, Aditya Mahadevan, Arvind Suresh, Yuta Ishiyama, Yiping Li, Julian C Hong, Autl J Butte

{"title":"From Zero-Shot to Bedside: A Practical Playbook for Adapting Open-Source Large Language Models to Clinical Symptom Extraction.","authors":"Li-Ching Chen, Travis Zack, Divneet Mandair, Aditya Mahadevan, Arvind Suresh, Yuta Ishiyama, Yiping Li, Julian C Hong, Autl J Butte","doi":"","DOIUrl":"","url":null,"abstract":"Large language models (LLMs) are increasingly applied to clinical notes, but guidance on how to adapt open-source models to specific tasks and manage annotation quality at scale is limited. We present a playbook for fine-tuning LLMs on de-identified clinical notes from patients with pancreatic cancer, spanning both pre-diagnosis and on-treatment settings. We evaluate prompting strategies, contrast open-source models with GPT-4o, and explore disease-level versus task-specific adaptation. A key contribution is an LLM-assisted adjudication workflow in which models flag notes where predictions consistently conflict with initial human labels. This approach concentrated expert review on a small fraction of cases while identifying many true annotation errors, ultimately improving downstream model performance. We further examine the use of machine-generated annotations to augment limited expert labels, showing that balanced mixtures of synthetic and human data can enhance fine-tuned models. Our findings provide practical guidance for deploying open-source LLMs in clinical contexts, offering strategies to improve accuracy, reduce annotation burden, and enable privacy-preserving, site-adapted clinical natural language processing (NLP).","PeriodicalId":74504,"journal":{"name":"Proceedings of machine learning research","volume":"297 ","pages":"1023-1046"},"PeriodicalIF":0.0,"publicationDate":"2026-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13345690/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148414262","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Censoring-Aware Reinforcement Learning to Optimize Early Risk Alerts from Longitudinal Clinical Data. 从纵向临床数据优化早期风险警报的审查意识强化学习。

Proceedings of machine learning research Pub Date : 2026-01-01

Qin Weng, Benjamin A Goldstein, Matthew M Engelhard

{"title":"Censoring-Aware Reinforcement Learning to Optimize Early Risk Alerts from Longitudinal Clinical Data.","authors":"Qin Weng, Benjamin A Goldstein, Matthew M Engelhard","doi":"","DOIUrl":"","url":null,"abstract":"Early recognition of chronic conditions is critical to ensure patients receive timely interventions and support. Passive surveillance of routine electronic health records (EHRs) provides information about longitudinal health trajectories that can support prompt recognition and inform associated early actions. However, relevant information is acquired at a different rate for each patient, and there is an inherent trade-off between the earliness versus the specificity of diagnosis and related actions. Therefore, determining when to alert providers about a likely chronic condition requires us to weigh the predicted risk at the given time against the anticipated value of future information. To address this challenge, we analyze the optimal timing of early alerts using a Partially Observable Markov Decision Process (POMDP) with asymmetric reward. To learn an optimal alerting policy, we then propose a model-free reinforcement learning (RL) framework tailored to long-term clinical event surveillance from EHRs. Our proposed framework overcomes the pervasive issue of right-censoring in offline EHRs by leveraging a pseudo-label imputation approach. We also analytically demonstrate that entropy-regularized RL enables post-hoc threshold calibration to adapt the learned policy to specific preferences regarding the importance of earliness versus specificity without retraining. Systematic evaluations in synthetic data reveal that the advantage of RL-based look-ahead planning is maximized when diagnostic evidence emerges in predictable information bursts. Finally, real-world validations on two clinical cohorts show that our policy achieves an actionable lead time of 20.9 months prior to Alzheimer's disease diagnosis and 8.7 months prior to autism diagnosis in a pediatric cohort, both at 90% specificity. Our method and results provide a generalizable blueprint for optimal surveillance of chronic disease processes.","PeriodicalId":74504,"journal":{"name":"Proceedings of machine learning research","volume":"340 ","pages":"2151-2174"},"PeriodicalIF":0.0,"publicationDate":"2026-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13626436/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149119683","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0