Searcharxiv⌕ Search

arXiv subjects

Eric Karl Oermann

Publications and source records attributed to Eric Karl Oermann.

At least 19 recordsLinked to original sources

Discharge Summary Generation Is Not Just Summarization: A Reality Check for Evaluation

Discharge summaries combine a retrospective account of hospitalization with discharge-time decisions. Physician-written references therefore contain both facts supported by prior clinical notes and reference-only content, potentially reflecting clinical synthesis, handoffs, or discharge planning. Reference-only evaluation can conflate failure to recover available information with failure to reproduce content absent from the model input. We introduce FairSumEval, an evidence-aware, fact-level framework that assesses whether each reference and generated fact is supported by the model's evidence and captured by the other summary. The resulting outcome map separates matches, omissions, and additions. Two complementary metrics, Soft IoU-S and Soft IoU-NS, measure reference-generation agreement for facts with and without identified evidence support. On 200 MIMIC-III admissions, 18.8% of physician reference facts have identified support in the prior evidence, compared with 65.4% and 63.5% of facts generated by a general-purpose and a medically specialized model. Overlap is substantially higher for supported content (0.255 and 0.236) than for content without identified support (0.030 and 0.026). The two scores are weakly correlated across admissions (rho = 0.22 and 0.17), and conventional metrics do not consistently capture both dimensions. FairSumEval distinguishes evidence-grounded summarization from agreement on content without identified support, revealing what models omit and add in each category.

cs.CL↗

Clinician use of language models diverges from how the models are evaluated

Large language model (LLM) assistants are being deployed to clinicians across health systems, and judgments about their readiness rest largely on benchmark scores, most of them derived from examination questions or curated cases. A benchmark predicts performance in deployment only to the extent that its items resemble real use, yet whether benchmarks reflect the work these systems receive has rarely been measured. Here we analyze 127,833 queries sent by 6,342 physicians, advanced practice providers and nurses in 35 specialties to an institutional assistant during an eight-month roll-out. We characterize each query with RCQ-Map, a clinician-validated framework grounded in taxonomies of clinical questions and of LLM evaluation, which records its task, intent, answerability, missing information and potential harm. Documentation and administration (36.2%) and knowledge retrieval (28.9%) made up nearly two-thirds of use, and diagnosis 3.7%; more than a third of queries could not be answered well as posed. Applying RCQ-Map to 58 public benchmarks drawn from major evaluation suites and frontier model reports, which we assemble into the Clinical AI Benchmark Atlas, showed that the median benchmark contained no documentation requests and shared 31% of the task mix of real use, less than an even spread across task categories would. Benchmarks in suites designed to resemble clinical practice were individually no closer to real use than those used in frontier model reports. Benchmark scores therefore say little about how clinical AI performs on most of the work it is actually given, and evaluation should be matched to real clinical use.

cs.CL↗

Reason in Style: Discovering and Controlling Style in Language Models

Language models learn content and style jointly, making stylistic variation in their outputs difficult to identify and control. We study whether recurring styles in model responses can be discovered without supervision and explicitly controlled. We design an algorithm that learns to separate representations of content and style from language models' outputs and validate its effectiveness on math questions in a controlled setting. By applying this method to over 100K verified traces from nine distinct teacher models, we discover six recurring yet imbalanced styles. We then fine-tune smaller student models to follow these styles when explicitly conditioned on them, using importance weighting to balance the contribution of the styles represented in the corpus. This approach improves Pass@$k$ over standard fine-tuning on the same data across six math reasoning benchmarks, demonstrating that we can diversify the style of answers effectively. We confirm that this also results in strong correspondence between requested and realized styles. We find that style affects correctness: the probability of solving a problem depends on the style we condition on, and different problems benefit from different styles. In summary, our results show that stylistic variation in model-generated data can be discovered in an unsupervised way, and made explicit, providing a source of both control and improved reasoning performance.

cs.CL↗

Multi-Task Bayesian In-Context Learning

Bayesian predictive inference provides a principled framework for uncertainty quantification, data efficiency, and robust generalization. However, exact inference is often intractable, and scalable approximations may remain computationally expensive or require restrictive modeling assumptions that degrade predictive performance. Prior-Data Fitted and in-context models have recently emerged as an amortized alternative by learning to map datasets directly to predictive distributions, but existing approaches are tightly coupled to the support of the training prior and lack explicit mechanisms for adapting to new priors at test time, resulting in limited robustness under distribution shift. We introduce a multi-task in-context learning framework for amortized hierarchical Bayesian predictive inference that explicitly represents prior information as a prefix of in-context datasets. A transformer trained on sequences of prior and target tasks learns to adapt its predictions across families of priors. On a suite of evaluations with increasing difficulty, including out-of-meta-distribution priors and priors with high-dimensional latent structures, our method matches oracle Bayesian predictors while being orders of magnitude faster. We further demonstrate its practical relevance on a real-world spatiotemporal temperature prediction benchmark. Code is available at https://github.com/martianmartina/multi-task-bayesian-icl/.

cs.LG↗

Large Language Models Predict Functional Outcomes after Acute Ischemic Stroke

Accurate prediction of functional outcomes after acute ischemic stroke can inform clinical decision-making and resource allocation. Prior work on modified Rankin Scale (mRS) prediction has relied primarily on structured variables (e.g., age, NIHSS) and conventional machine learning. The ability of large language models (LLMs) to infer future mRS scores directly from routine admission notes remains largely unexplored. We evaluated encoder (BERT, NYUTron) and generative (Llama-3.1-8B, MedGemma-4B) LLMs, in both frozen and fine-tuned settings, for discharge and 90-day mRS prediction using a large, real-world stroke registry. The discharge outcome dataset included 9,485 History and Physical notes and the 90-day outcome dataset included 1,898 notes from the NYU Langone Get With The Guidelines-Stroke registry (2016-2025). Data were temporally split with the most recent 12 months held out for testing. Performance was assessed using exact (7-class) mRS accuracy and binary functional outcome (mRS 0-2 vs. 3-6) accuracy and compared against established structured-data baselines incorporating NIHSS and age. Fine-tuned Llama achieved the highest performance, with 90-day exact mRS accuracy of 33.9% [95% CI, 27.9-39.9%] and binary accuracy of 76.3% [95% CI, 70.7-81.9%]. Discharge performance reached 42.0% [95% CI, 39.0-45.0%] exact accuracy and 75.0% [95% CI, 72.4-77.6%] binary accuracy. For 90-day prediction, Llama performed comparably to structured-data baselines. Fine-tuned LLMs can predict post-stroke functional outcomes from admission notes alone, achieving performance comparable to models requiring structured variable abstraction. Our findings support the development of text-based prognostic tools that integrate seamlessly into clinical workflows without manual data extraction.

cs.LG↗

Generalist Large Language Models Outperform Clinical Tools on Medical Benchmarks

Specialized clinical AI assistants are rapidly entering medical practice, often framed as safer or more reliable than general-purpose large language models (LLMs). Yet, unlike frontier models, these clinical tools are rarely subjected to independent, quantitative evaluation, creating a critical evidence gap despite their growing influence on diagnosis, triage, and guideline interpretation. We assessed two widely deployed clinical AI systems (OpenEvidence and UpToDate Expert AI) against three state-of-the-art generalist LLMs (GPT-5, Gemini 3 Pro, and Claude Sonnet 4.5) using a 1,000-item mini-benchmark combining MedQA (medical knowledge) and HealthBench (clinician-alignment) tasks. Generalist models consistently outperformed clinical tools, with GPT-5 achieving the highest scores, while OpenEvidence and UpToDate demonstrated deficits in completeness, communication quality, context awareness, and systems-based safety reasoning. These findings reveal that tools marketed for clinical decision support may often lag behind frontier LLMs, underscoring the urgent need for transparent, independent evaluation before deployment in patient-facing workflows.

cs.CL↗

CNS-Obsidian: A Neurosurgical Vision-Language Model Built From Scientific Publications

General-purpose VLMs demonstrate impressive capabilities, but their opaque training on uncurated internet data poses critical limitations for high-stakes decision-making, such as in neurosurgery. We present CNS-Obsidian, a neurosurgical VLM trained on peer-reviewed literature, and demonstrate its clinical utility versus GPT-4o in a real-world setting. We compiled 23,984 articles from Neurosurgery Publications journals, yielding 78,853 figures and captions. Using GPT-4o and Claude Sonnet-3.5, we converted these into 263,064 training samples across three formats: instruction fine-tuning, multiple-choice questions, and differential diagnosis. We trained CNS-Obsidian, a fine-tune of the 34-billion parameter LLaVA-Next model. In a blinded, randomized trial at NYU Langone Health (Aug 30-Nov 30, 2024), neurosurgery consultations were assigned to either CNS-Obsidian or a HIPAA-compliant GPT-4o endpoint as diagnostic co-pilot after consultations. Primary outcomes were diagnostic helpfulness and accuracy, assessed via user ratings and presence of correct diagnosis within the VLM-provided differential. CNS-Obsidian matched GPT-4o on synthetic questions (76.13% vs 77.54%, p=0.235), but only achieved 46.81% accuracy on human-generated questions versus GPT-4o's 65.70% (p<10-15). In the randomized trial, 70 consultations were evaluated (32 CNS-Obsidian, 38 GPT-4o) from 959 total consults (7.3% utilization). CNS-Obsidian received positive ratings in 40.62% of cases versus 57.89% for GPT-4o (p=0.230). Both models included correct diagnosis in approximately 60% of cases (59.38% vs 65.79%, p=0.626). Domain-specific VLMs trained on curated scientific literature can approach frontier model performance despite being orders of magnitude smaller and less expensive to train. This establishes a transparent framework for scientific communities to build specialized AI models.

cs.AI↗

Generalist Foundation Models Are Not Clinical Enough for Hospital Operations

Hospitals and healthcare systems rely on operational decisions that determine patient flow, cost, and quality of care. Despite strong performance on medical knowledge and conversational benchmarks, foundation models trained on general text may lack the specialized knowledge required for these operational decisions. We introduce Lang1, a family of models (100M-7B parameters) pretrained on a specialized corpus blending 80B clinical tokens from NYU Langone Health's EHRs and 627B tokens from the internet. To rigorously evaluate Lang1 in real-world settings, we developed the REalistic Medical Evaluation (ReMedE), a benchmark derived from 668,331 EHR notes that evaluates five critical tasks: 30-day readmission prediction, 30-day mortality prediction, length of stay, comorbidity coding, and predicting insurance claims denial. In zero-shot settings, both general-purpose and specialized models underperform on four of five tasks (36.6%-71.7% AUROC), with mortality prediction being an exception. After finetuning, Lang1-1B outperforms finetuned generalist models up to 70x larger and zero-shot models up to 671x larger, improving AUROC by 3.64%-6.75% and 1.66%-23.66% respectively. We also observed cross-task scaling with joint finetuning on multiple tasks leading to improvement on other tasks. Lang1-1B effectively transfers to out-of-distribution settings, including other clinical tasks and an external health system. Our findings suggest that predictive capabilities for hospital operations require explicit supervised finetuning, and that this finetuning process is made more efficient by in-domain pretraining on EHR. Our findings support the emerging view that specialized LLMs can compete with generalist models in specialized tasks, and show that effective healthcare systems AI requires the combination of in-domain pretraining, supervised finetuning, and real-world evaluation beyond proxy benchmarks.

cs.CL↗

MedMobile: A mobile-sized language model with clinical capabilities

Language models (LMs) have demonstrated expert-level reasoning and recall abilities in medicine. However, computational costs and privacy concerns are mounting barriers to wide-scale implementation. To address these significant limitations, we introduce a parsimonious adaptation of phi-3-mini, MedMobile, a 3.8 billion parameter LM capable of running on a mobile device, for medical applications. We perform a careful set of pipeline additions and demonstrate that chain of thought, ensembling, and fine-tuning lead to the greatest performance gains, while unexpectedly retrieval augmented generation fails to demonstrate significant improvements. We evaluate the efficiency of our pipeline on the MultiMedQA and MedBullets. We demonstrate that MedMobile scores 75.7% on the MedQA (USMLE), surpassing the passing mark for licensed physicians (~60%) and rivaling scores of models 100 times its size. Across the entirety of the MultiMedQA, MedMobile achieves SOTA performance for models with less than 5B parameters and represents the smallest model to pass the MedQA (USMLE). MedMobile holds promise to democratize access to language models in medicine, bolstering lower compute needs and fast inference speeds. With the ability to combat the biggest barriers to entry for language models in medicine, we hope that MedMobile is a critical step forward in developing clinically relevant language models.

cs.CL↗

On the Relationship Between the Choice of Representation and In-Context Learning

In-context learning (ICL) is the ability of a large language model (LLM) to learn a new task from a few demonstrations presented as part of the context. Past studies have attributed a large portion of the success of ICL to the way these in-context demonstrations are represented, particularly to how labels are represented in classification tasks. On the other hand, observations of the learning capacity of ICL (i.e., the extent to which more in-context demonstrations can lead to higher performance) have been mixed, and ICL is often thought to occur only under specific conditions. The interaction between these two aspects in ICL, representation and learning, has not been studied in depth until now. We hypothesize that they are largely independent of one another, such that the representation of demonstrations determines the baseline accuracy of ICL, while learning from additional demonstrations improves only on top of this baseline. We validate this hypothesis by developing an optimization algorithm that can enumerate a spectrum of possible label sets (representations) varying in semantic relevance. We then perform ICL with varying numbers of in-context demonstrations for each of these label sets. We observed that learning happens regardless of the quality of the label set itself, although its efficiency, measured by the slope of improvement over in-context demonstrations, is conditioned on both the label set quality and the parameter count of the underlying language model. Despite the emergence of learning, the relative quality (accuracy) of the choice of a label set (representation) is largely maintained throughout learning, confirming our hypothesis and implying their orthogonality. Our work reveals a previously underexplored aspect of ICL: the independent effects of learning from demonstrations and their representations on ICL performance.

cs.CL↗

Clinically Grounded Agent-based Report Evaluation: An Interpretable Metric for Radiology Report Generation

Radiological imaging is central to diagnosis, treatment planning, and clinical decision-making. Vision-language foundation models have spurred interest in automated radiology report generation (RRG), but safe deployment requires reliable clinical evaluation of generated reports. Existing metrics often rely on surface-level similarity or behave as black boxes, lacking interpretability. We introduce ICARE (Interpretable and Clinically-grounded Agent-based Report Evaluation), an interpretable evaluation framework leveraging large language model agents and dynamic multiple-choice question answering (MCQA). Two agents, each with either the ground-truth or generated report, generate clinically meaningful questions and quiz each other. Agreement on answers captures preservation and consistency of findings, serving as interpretable proxies for clinical precision and recall. By linking scores to question-answer pairs, ICARE enables transparent, and interpretable assessment. Clinician studies show ICARE aligns significantly more with expert judgment than prior metrics. Perturbation analyses confirm sensitivity to clinical content and reproducibility, while model comparisons reveal interpretable error patterns.

cs.CL↗

Evaluating the performance and fragility of large language models on the self-assessment for neurological surgeons

The Congress of Neurological Surgeons Self-Assessment for Neurological Surgeons (CNS-SANS) questions are widely used by neurosurgical residents to prepare for written board examinations. Recently, these questions have also served as benchmarks for evaluating large language models' (LLMs) neurosurgical knowledge. This study aims to assess the performance of state-of-the-art LLMs on neurosurgery board-like questions and to evaluate their robustness to the inclusion of distractor statements. A comprehensive evaluation was conducted using 28 large language models. These models were tested on 2,904 neurosurgery board examination questions derived from the CNS-SANS. Additionally, the study introduced a distraction framework to assess the fragility of these models. The framework incorporated simple, irrelevant distractor statements containing polysemous words with clinical meanings used in non-clinical contexts to determine the extent to which such distractions degrade model performance on standard medical benchmarks. 6 of the 28 tested LLMs achieved board-passing outcomes, with the top-performing models scoring over 15.7% above the passing threshold. When exposed to distractions, accuracy across various model architectures was significantly reduced-by as much as 20.4%-with one model failing that had previously passed. Both general-purpose and medical open-source models experienced greater performance declines compared to proprietary variants when subjected to the added distractors. While current LLMs demonstrate an impressive ability to answer neurosurgery board-like exam questions, their performance is markedly vulnerable to extraneous, distracting information. These findings underscore the critical need for developing novel mitigation strategies aimed at bolstering LLM resilience against in-text distractions, particularly for safe and effective clinical deployment.

cs.CL↗

Medical large language models are easily distracted

Large language models (LLMs) have the potential to transform medicine, but real-world clinical scenarios contain extraneous information that can hinder performance. The rise of assistive technologies like ambient dictation, which automatically generates draft notes from live patient encounters, has the potential to introduce additional noise making it crucial to assess the ability of LLM's to filter relevant data. To investigate this, we developed MedDistractQA, a benchmark using USMLE-style questions embedded with simulated real-world distractions. Our findings show that distracting statements (polysemous words with clinical meanings used in a non-clinical context or references to unrelated health conditions) can reduce LLM accuracy by up to 17.9%. Commonly proposed solutions to improve model performance such as retrieval-augmented generation (RAG) and medical fine-tuning did not change this effect and in some cases introduced their own confounders and further degraded performance. Our findings suggest that LLMs natively lack the logical mechanisms necessary to distinguish relevant from irrelevant clinical information, posing challenges for real-world applications. MedDistractQA and our results highlights the need for robust mitigation strategies to enhance LLM resilience to extraneous information.

cs.CL↗

It is Too Many Options: Pitfalls of Multiple-Choice Questions in Generative AI and Medical Education

The performance of Large Language Models (LLMs) on multiple-choice question (MCQ) benchmarks is frequently cited as proof of their medical capabilities. We hypothesized that LLM performance on medical MCQs may in part be illusory and driven by factors beyond medical content knowledge and reasoning capabilities. To assess this, we created a novel benchmark of free-response questions with paired MCQs (FreeMedQA). Using this benchmark, we evaluated three state-of-the-art LLMs (GPT-4o, GPT-3.5, and LLama-3-70B-instruct) and found an average absolute deterioration of 39.43% in performance on free-response questions relative to multiple-choice (p = 1.3 * 10-5) which was greater than the human performance decline of 22.29%. To isolate the role of the MCQ format on performance, we performed a masking study, iteratively masking out parts of the question stem. At 100% masking, the average LLM multiple-choice performance was 6.70% greater than random chance (p = 0.002) with one LLM (GPT-4o) obtaining an accuracy of 37.34%. Notably, for all LLMs the free-response performance was near zero. Our results highlight the shortcomings in medical MCQ benchmarks for overestimating the capabilities of LLMs in medicine, and, broadly, the potential for improving both human and machine assessments using LLM-evaluated free-response questions.

cs.CL↗

BPQA Dataset: Evaluating How Well Language Models Leverage Blood Pressures to Answer Biomedical Questions

Clinical measurements such as blood pressures and respiration rates are critical in diagnosing and monitoring patient outcomes. It is an important component of biomedical data, which can be used to train transformer-based language models (LMs) for improving healthcare delivery. It is, however, unclear whether LMs can effectively interpret and use clinical measurements. We investigate two questions: First, can LMs effectively leverage clinical measurements to answer related medical questions? Second, how to enhance an LM's performance on medical question-answering (QA) tasks that involve measurements? We performed a case study on blood pressure readings (BPs), a vital sign routinely monitored by medical professionals. We evaluated the performance of four LMs: BERT, BioBERT, MedAlpaca, and GPT-3.5, on our newly developed dataset, BPQA (Blood Pressure Question Answering). BPQA contains $100$ medical QA pairs that were verified by medical students and designed to rely on BPs . We found that GPT-3.5 and MedAlpaca (larger and medium sized LMs) benefit more from the inclusion of BPs than BERT and BioBERT (small sized LMs). Further, augmenting measurements with labels improves the performance of BioBERT and Medalpaca (domain specific LMs), suggesting that retrieval may be useful for improving domain-specific LMs.

cs.CL↗

MedG-KRP: Medical Graph Knowledge Representation Probing

Large language models (LLMs) have recently emerged as powerful tools, finding many medical applications. LLMs' ability to coalesce vast amounts of information from many sources to generate a response-a process similar to that of a human expert-has led many to see potential in deploying LLMs for clinical use. However, medicine is a setting where accurate reasoning is paramount. Many researchers are questioning the effectiveness of multiple choice question answering (MCQA) benchmarks, frequently used to test LLMs. Researchers and clinicians alike must have complete confidence in LLMs' abilities for them to be deployed in a medical setting. To address this need for understanding, we introduce a knowledge graph (KG)-based method to evaluate the biomedical reasoning abilities of LLMs. Essentially, we map how LLMs link medical concepts in order to better understand how they reason. We test GPT-4, Llama3-70b, and PalmyraMed-70b, a specialized medical model. We enlist a panel of medical students to review a total of 60 LLM-generated graphs and compare these graphs to BIOS, a large biomedical KG. We observe GPT-4 to perform best in our human review but worst in our ground truth comparison; vice-versa with PalmyraMed, the medical model. Our work provides a means of visualizing the medical reasoning pathways of LLMs so they can be implemented in clinical settings safely and effectively.

cs.AI↗

Refining Packing and Shuffling Strategies for Enhanced Performance in Generative Language Models

Packing and shuffling tokens is a common practice in training auto-regressive language models (LMs) to prevent overfitting and improve efficiency. Typically documents are concatenated to chunks of maximum sequence length (MSL) and then shuffled. However setting the atom size, the length for each data chunk accompanied by random shuffling, to MSL may lead to contextual incoherence due to tokens from different documents being packed into the same chunk. An alternative approach is to utilize padding, another common data packing strategy, to avoid contextual incoherence by only including one document in each shuffled chunk. To optimize both packing strategies (concatenation vs padding), we investigated the optimal atom size for shuffling and compared their performance and efficiency. We found that matching atom size to MSL optimizes performance for both packing methods (concatenation and padding), and padding yields lower final perplexity (higher performance) than concatenation at the cost of more training steps and lower compute efficiency. This trade-off informs the choice of packing methods in training language models.

cs.CL↗

Generalization in Healthcare AI: Evaluation of a Clinical Large Language Model

Advances in large language models (LLMs) provide new opportunities in healthcare for improved patient care, clinical decision-making, and enhancement of physician and administrator workflows. However, the potential of these models importantly depends on their ability to generalize effectively across clinical environments and populations, a challenge often underestimated in early development. To better understand reasons for these challenges and inform mitigation approaches, we evaluated ClinicLLM, an LLM trained on [HOSPITAL]'s clinical notes, analyzing its performance on 30-day all-cause readmission prediction focusing on variability across hospitals and patient characteristics. We found poorer generalization particularly in hospitals with fewer samples, among patients with government and unspecified insurance, the elderly, and those with high comorbidities. To understand reasons for lack of generalization, we investigated sample sizes for fine-tuning, note content (number of words per note), patient characteristics (comorbidity level, age, insurance type, borough), and health system aspects (hospital, all-cause 30-day readmission, and mortality rates). We used descriptive statistics and supervised classification to identify features. We found that, along with sample size, patient age, number of comorbidities, and the number of words in notes are all important factors related to generalization. Finally, we compared local fine-tuning (hospital specific), instance-based augmented fine-tuning and cluster-based fine-tuning for improving generalization. Among these, local fine-tuning proved most effective, increasing AUC by 0.25% to 11.74% (most helpful in settings with limited data). Overall, this study provides new insights for enhancing the deployment of large language models in the societally important domain of healthcare, and improving their performance for broader populations.

cs.CL↗