arXiv subjects

Mihaela van der Schaar

Publications and source records attributed to Mihaela van der Schaar.

At least 55 records · Page 3Linked to original sources

arXiv2026

The Reasoning-Creativity Trade-off: Toward Creativity-Driven Problem Solving

State-of-the-art large language model (LLM) pipelines rely on bootstrapped reasoning loops: sampling diverse chains of thought and reinforcing the highest-scoring ones, mainly optimizing correctness. We analyze how this design choice is sensitive to the collapse of the model's distribution over reasoning paths, slashing semantic entropy and undermining creative problem-solving. To analyze this failure, we introduce Distributional Creative Reasoning (DCR), a unified variational objective that casts training as gradient flow through probability measures on solution traces. STaR, GRPO, and DPO, as well as entropy bonuses, and other methods, all constitute special cases of the same loss. The framework delivers three core results: (i) the diversity decay theorem, describing how correctness-based objectives lead to distinct modes of diversity decay for STaR, GRPO, and DPO; (ii) designs that ensure convergence to a stable and diverse policy, effectively preventing collapse; and (iii) simple, actionable recipes to achieve this in practice. DCR thus offers the first principled recipe for LLMs that remain both correct and creative.

cs.LG↗

arXiv2025

Towards Human-Guided, Data-Centric LLM Co-Pilots

Machine learning (ML) has the potential to revolutionize various domains, but its adoption is often hindered by the disconnect between the needs of domain experts and translating these needs into robust and valid ML tools. Despite recent advances in LLM-based co-pilots to democratize ML for non-technical domain experts, these systems remain predominantly focused on model-centric aspects while overlooking critical data-centric challenges. This limitation is problematic in complex real-world settings where raw data often contains complex issues, such as missing values, label noise, and domain-specific nuances requiring tailored handling. To address this we introduce CliMB-DC, a human-guided, data-centric framework for LLM co-pilots that combines advanced data-centric tools with LLM-driven reasoning to enable robust, context-aware data processing. At its core, CliMB-DC introduces a novel, multi-agent reasoning system that combines a strategic coordinator for dynamic planning and adaptation with a specialized worker agent for precise execution. Domain expertise is then systematically incorporated to guide the reasoning process using a human-in-the-loop approach. To guide development, we formalize a taxonomy of key data-centric challenges that co-pilots must address. Thereafter, to address the dimensions of the taxonomy, we integrate state-of-the-art data-centric tools into an extensible, open-source architecture, facilitating the addition of new tools from the research community. Empirically, using real-world healthcare datasets we demonstrate CliMB-DC's ability to transform uncurated datasets into ML-ready formats, significantly outperforming existing co-pilot baselines for handling data-centric challenges. CliMB-DC promises to empower domain experts from diverse domains -- healthcare, finance, social sciences and more -- to actively participate in driving real-world impact using ML.

cs.LG↗

arXiv2025

Visualizing token importance for black-box language models

We consider the problem of auditing black-box large language models (LLMs) to ensure they behave reliably when deployed in production settings, particularly in high-stakes domains such as legal, medical, and regulatory compliance. Existing approaches for LLM auditing often focus on isolated aspects of model behavior, such as detecting specific biases or evaluating fairness. We are interested in a more general question -- can we understand how the outputs of black-box LLMs depend on each input token? There is a critical need to have such tools in real-world applications that rely on inaccessible API endpoints to language models. However, this is a highly non-trivial problem, as LLMs are stochastic functions (i.e. two outputs will be different by chance), while computing prompt-level gradients to approximate input sensitivity is infeasible. To address this, we propose Distribution-Based Sensitivity Analysis (DBSA), a lightweight model-agnostic procedure to evaluate the sensitivity of the output of a language model for each input token, without making any distributional assumptions about the LLM. DBSA is developed as a practical tool for practitioners, enabling quick, plug-and-play visual exploration of LLMs reliance on specific input tokens. Through illustrative examples, we demonstrate how DBSA can enable users to inspect LLM inputs and find sensitivities that may be overlooked by existing LLM interpretability methods.

cs.CL↗

arXiv2025

Semantic-KG: Using Knowledge Graphs to Construct Benchmarks for Measuring Semantic Similarity

Evaluating the open-form textual responses generated by Large Language Models (LLMs) typically requires measuring the semantic similarity of the response to a (human generated) reference. However, there is evidence that current semantic similarity methods may capture syntactic or lexical forms over semantic content. While benchmarks exist for semantic equivalence, they often suffer from high generation costs due to reliance on subjective human judgment, limited availability for domain-specific applications, and unclear definitions of equivalence. This paper introduces a novel method for generating benchmarks to evaluate semantic similarity methods for LLM outputs, specifically addressing these limitations. Our approach leverages knowledge graphs (KGs) to generate pairs of natural-language statements that are semantically similar or dissimilar, with dissimilar pairs categorized into one of four sub-types. We generate benchmark datasets in four different domains (general knowledge, biomedicine, finance, biology), and conduct a comparative study of semantic similarity methods including traditional natural language processing scores and LLM-as-a-judge predictions. We observe that the sub-type of semantic variation, as well as the domain of the benchmark impact the performance of semantic similarity methods, with no method being consistently superior. Our results present important implications for the use of LLM-as-a-judge in detecting the semantic content of text. Code is available at https://github.com/QiyaoWei/semantic-kg and the dataset is available at https://huggingface.co/datasets/QiyaoWei/Semantic-KG.

cs.AI↗

arXiv2025

What's the next frontier for Data-centric AI? Data Savvy Agents

The recent surge in AI agents that autonomously communicate, collaborate with humans and use diverse tools has unlocked promising opportunities in various real-world settings. However, a vital aspect remains underexplored: how agents handle data. Scalable autonomy demands agents that continuously acquire, process, and evolve their data. In this paper, we argue that data-savvy capabilities should be a top priority in the design of agentic systems to ensure reliable real-world deployment. Specifically, we propose four key capabilities to realize this vision: (1) Proactive data acquisition: enabling agents to autonomously gather task-critical knowledge or solicit human input to address data gaps; (2) Sophisticated data processing: requiring context-aware and flexible handling of diverse data challenges and inputs; (3) Interactive test data synthesis: shifting from static benchmarks to dynamically generated interactive test data for agent evaluation; and (4) Continual adaptation: empowering agents to iteratively refine their data and background knowledge to adapt to shifting environments. While current agent research predominantly emphasizes reasoning, we hope to inspire a reflection on the role of data-savvy agents as the next frontier in data-centric AI.

cs.LG↗

arXiv2025

Cascaded Language Models for Cost-effective Human-AI Decision-Making

A challenge in human-AI decision-making is to balance three factors: the correctness of predictions, the cost of knowledge and reasoning complexity, and the confidence about whether to abstain from automated answers or escalate to human experts. In this work, we present a cascaded LLM decision framework that adaptively delegates tasks across multiple tiers of expertise -- a base model for initial candidate answers, a more capable and knowledgeable (but costlier) large model, and a human expert for when the model cascade abstains. Our method proceeds in two stages. First, a deferral policy determines whether to accept the base model's answer or regenerate it with the large model based on the confidence score. Second, an abstention policy decides whether the cascade model response is sufficiently certain or requires human intervention. Moreover, to overcome static policies and accommodate changing task difficulty, we incorporate an online learning mechanism which uses human feedback. We demonstrate this approach to general question-answering (ARC-Easy, ARC-Challenge, and MMLU) and medical question-answering (MedQA and MedMCQA). Our results demonstrate that our cascaded strategy outperforms single-model baselines in most cases, achieving higher accuracy while reducing costs and providing a principled approach to handling abstentions.

cs.AI↗

arXiv2025

Improving the Generation and Evaluation of Synthetic Data for Downstream Medical Causal Inference

Causal inference is essential for developing and evaluating medical interventions, yet real-world medical datasets are often difficult to access due to regulatory barriers. This makes synthetic data a potentially valuable asset that enables these medical analyses, along with the development of new inference methods themselves. Generative models can produce synthetic data that closely approximate real data distributions, yet existing methods do not consider the unique challenges that downstream causal inference tasks, and specifically those focused on treatments, pose. We establish a set of desiderata that synthetic data containing treatments should satisfy to maximise downstream utility: preservation of (i) the covariate distribution, (ii) the treatment assignment mechanism, and (iii) the outcome generation mechanism. Based on these desiderata, we propose a set of evaluation metrics to assess such synthetic data. Finally, we present STEAM: a novel method for generating Synthetic data for Treatment Effect Analysis in Medicine that mimics the data-generating process of data containing treatments and optimises for our desiderata. We empirically demonstrate that STEAM achieves state-of-the-art performance across our metrics as compared to existing generative models, particularly as the complexity of the true data-generating process increases.

cs.LG↗

arXiv2025

Simulating Viva Voce Examinations to Evaluate Clinical Reasoning in Large Language Models

Clinical reasoning in medicine is a hypothesis-driven process where physicians refine diagnoses from limited information through targeted history, physical examination, and diagnostic investigations. In contrast, current medical benchmarks for large language models (LLMs) primarily assess knowledge recall through single-turn questions, where complete clinical information is provided upfront. To address this gap, we introduce VivaBench, a multi-turn benchmark that evaluates sequential clinical reasoning in LLM agents. Our dataset consists of 1762 physician-curated clinical vignettes structured as interactive scenarios that simulate a (oral) examination in medical training, requiring agents to actively probe for relevant findings, select appropriate investigations, and synthesize information across multiple steps to reach a diagnosis. While current LLMs demonstrate competence in diagnosing conditions from well-described clinical presentations, their performance degrades significantly when required to navigate iterative diagnostic reasoning under uncertainty in our evaluation. Our analysis identified several failure modes that mirror common cognitive errors in clinical practice, including: (1) fixation on initial hypotheses, (2) inappropriate investigation ordering, (3) premature diagnostic closure, and (4) failing to screen for critical conditions. These patterns reveal fundamental limitations in how current LLMs reason and make decisions under uncertainty. Through VivaBench, we provide a standardized benchmark for evaluating conversational medical AI systems for real-world clinical decision support. Beyond medical applications, we contribute to the larger corpus of research on agentic AI by demonstrating how sequential reasoning trajectories can diverge in complex decision-making environments.

cs.LG↗

arXiv2025

Hypothesis Hunting with Evolving Networks of Autonomous Scientific Agents

Large-scale scientific datasets -- spanning health biobanks, cell atlases, Earth reanalyses, and more -- create opportunities for exploratory discovery unconstrained by specific research questions. We term this process hypothesis hunting: the cumulative search for insight through sustained exploration across vast and complex hypothesis spaces. To support it, we introduce AScience, a framework modeling discovery as the interaction of agents, networks, and evaluation norms, and implement it as ASCollab, a distributed system of LLM-based research agents with heterogeneous behaviors. These agents self-organize into evolving networks, continually producing and peer-reviewing findings under shared standards of evaluation. Experiments show that such social dynamics enable the accumulation of expert-rated results along the diversity-quality-novelty frontier, including rediscoveries of established biomarkers, extensions of known pathways, and proposals of new therapeutic targets. While wet-lab validation remains indispensable, our experiments on cancer cohorts demonstrate that socially structured, agentic networks can sustain exploratory hypothesis hunting at scale.

cs.AI↗

arXiv2025

Technical Report: Facilitating the Adoption of Causal Inference Methods Through LLM-Empowered Co-Pilot

Estimating treatment effects (TE) from observational data is a critical yet complex task in many fields, from healthcare and economics to public policy. While recent advances in machine learning and causal inference have produced powerful estimation techniques, their adoption remains limited due to the need for deep expertise in causal assumptions, adjustment strategies, and model selection. In this paper, we introduce CATE-B, an open-source co-pilot system that uses large language models (LLMs) within an agentic framework to guide users through the end-to-end process of treatment effect estimation. CATE-B assists in (i) constructing a structural causal model via causal discovery and LLM-based edge orientation, (ii) identifying robust adjustment sets through a novel Minimal Uncertainty Adjustment Set criterion, and (iii) selecting appropriate regression methods tailored to the causal structure and dataset characteristics. To encourage reproducibility and evaluation, we release a suite of benchmark tasks spanning diverse domains and causal complexities. By combining causal inference with intelligent, interactive assistance, CATE-B lowers the barrier to rigorous causal analysis and lays the foundation for a new class of benchmarks in automated treatment effect estimation.

cs.LG↗

arXiv2025

Stochastic Encodings for Active Feature Acquisition

Active Feature Acquisition is an instance-wise, sequential decision making problem. The aim is to dynamically select which feature to measure based on current observations, independently for each test instance. Common approaches either use Reinforcement Learning, which experiences training difficulties, or greedily maximize the conditional mutual information of the label and unobserved features, which makes myopic acquisitions. To address these shortcomings, we introduce a latent variable model, trained in a supervised manner. Acquisitions are made by reasoning about the features across many possible unobserved realizations in a stochastic latent space. Extensive evaluation on a large range of synthetic and real datasets demonstrates that our approach reliably outperforms a diverse set of baselines.

cs.LG↗

arXiv2025

Continuously Updating Digital Twins using Large Language Models

Digital twins are models of real-world systems that can simulate their dynamics in response to potential actions. In complex settings, the state and action variables, and available data and knowledge relevant to a system can constantly change, requiring digital twins to continuously update with these changes to remain relevant. Current approaches struggle in this regard, as they require fixed, well-defined modelling environments, and they cannot adapt to novel variables without re-designs, or incorporate new information without re-training. To address this, we frame digital twinning as an in-context learning problem using large language models, enabling seamless updates to the twin at inference time. We develop CALM-DT, a Context-Adaptive Language Model-based Digital Twin that can accurately simulate across diverse state-action spaces using in-context learning alone by utilising fine-tuned encoders for sample retrieval. We empirically demonstrate CALM-DT's competitive performance with existing digital twin approaches, and its unique ability to adapt to changes in its modelling environment without parameter updates.

cs.CL↗

arXiv2025

Beyond the ATE: Interpretable Modelling of Treatment Effects over Dose and Time

The Average Treatment Effect (ATE) is a foundational metric in causal inference, widely used to assess intervention efficacy in randomized controlled trials (RCTs). However, in many applications -- particularly in healthcare -- this static summary fails to capture the nuanced dynamics of treatment effects that vary with both dose and time. We propose a framework for modelling treatment effect trajectories as smooth surfaces over dose and time, enabling the extraction of clinically actionable insights such as onset time, peak effect, and duration of benefit. To ensure interpretability, robustness, and verifiability -- key requirements in high-stakes domains -- we adapt SemanticODE, a recent framework for interpretable trajectory modelling, to the causal setting where treatment effects are never directly observed. Our approach decouples the estimation of trajectory shape from the specification of clinically relevant properties (e.g., maxima, inflection points), supporting domain-informed priors, post-hoc editing, and transparent analysis. We show that our method yields accurate, interpretable, and editable models of treatment dynamics, facilitating both rigorous causal analysis and practical decision-making.

cs.LG↗

arXiv2025

Inverse Reinforcement Learning Meets Large Language Model Post-Training: Basics, Advances, and Opportunities

In the era of Large Language Models (LLMs), alignment has emerged as a fundamental yet challenging problem in the pursuit of more reliable, controllable, and capable machine intelligence. The recent success of reasoning models and conversational AI systems has underscored the critical role of reinforcement learning (RL) in enhancing these systems, driving increased research interest at the intersection of RL and LLM alignment. This paper provides a comprehensive review of recent advances in LLM alignment through the lens of inverse reinforcement learning (IRL), emphasizing the distinctions between RL techniques employed in LLM alignment and those in conventional RL tasks. In particular, we highlight the necessity of constructing neural reward models from human data and discuss the formal and practical implications of this paradigm shift. We begin by introducing fundamental concepts in RL to provide a foundation for readers unfamiliar with the field. We then examine recent advances in this research agenda, discussing key challenges and opportunities in conducting IRL for LLM alignment. Beyond methodological considerations, we explore practical aspects, including datasets, benchmarks, evaluation metrics, infrastructure, and computationally efficient training and inference techniques. Finally, we draw insights from the literature on sparse-reward RL to identify open questions and potential research directions. By synthesizing findings from diverse studies, we aim to provide a structured and critical overview of the field, highlight unresolved challenges, and outline promising future directions for improving LLM alignment through RL and IRL techniques.

cs.LG↗

arXiv2025

Not All Explanations for Deep Learning Phenomena Are Equally Valuable

Developing a better understanding of surprising or counterintuitive phenomena has constituted a significant portion of deep learning research in recent years. These include double descent, grokking, and the lottery ticket hypothesis -- among many others. Works in this area often develop ad hoc hypotheses attempting to explain these observed phenomena on an isolated, case-by-case basis. This position paper asserts that, in many prominent cases, there is little evidence to suggest that these phenomena appear in real-world applications and these efforts may be inefficient in driving progress in the broader field. Consequently, we argue against viewing them as isolated puzzles that require bespoke resolutions or explanations. However, despite this, we suggest that deep learning phenomena do still offer research value by providing unique settings in which we can refine our broad explanatory theories of more general deep learning principles. This position is reinforced by analyzing the research outcomes of several prominent examples of these phenomena from the recent literature. We revisit the current norms in the research community in approaching these problems and propose practical recommendations for future research, aiming to ensure that progress on deep learning phenomena is well aligned with the ultimate pragmatic goal of progress in the broader field of deep learning.

cs.LG↗

arXiv2025

L2MAC: Large Language Model Automatic Computer for Extensive Code Generation

Transformer-based large language models (LLMs) are constrained by the fixed context window of the underlying transformer architecture, hindering their ability to produce long and coherent outputs. Memory-augmented LLMs are a promising solution, but current approaches cannot handle long output generation tasks since they (1) only focus on reading memory and reduce its evolution to the concatenation of new memories or (2) use very specialized memories that cannot adapt to other domains. This paper presents L2MAC, the first practical LLM-based general-purpose stored-program automatic computer (von Neumann architecture) framework, an LLM-based multi-agent system, for long and consistent output generation. Its memory has two components: the instruction registry, which is populated with a prompt program to solve the user-given task, and a file store, which will contain the final and intermediate outputs. Each instruction in turn is executed by a separate LLM agent, whose context is managed by a control unit capable of precise memory reading and writing to ensure effective interaction with the file store. These components enable L2MAC to generate extensive outputs, bypassing the constraints of the finite context window while producing outputs that fulfill a complex user-specified task. We empirically demonstrate that L2MAC achieves state-of-the-art performance in generating large codebases for system design tasks, significantly outperforming other coding methods in implementing the detailed user-specified task; we show that L2MAC works for general-purpose extensive text-based tasks, such as writing an entire book; and we provide valuable insights into L2MAC's performance improvement over existing methods.

cs.SE↗

arXiv2025

Revolutionizing Clinical Trials: A Manifesto for AI-Driven Transformation

This manifesto represents a collaborative vision forged by leaders in pharmaceuticals, consulting firms, clinical research, and AI. It outlines a roadmap for two AI technologies - causal inference and digital twins - to transform clinical trials, delivering faster, safer, and more personalized outcomes for patients. By focusing on actionable integration within existing regulatory frameworks, we propose a way forward to revolutionize clinical research and redefine the gold standard for clinical trials using AI.

cs.CY↗

arXiv2025

G-Sim: Generative Simulations with Large Language Models and Gradient-Free Calibration

Constructing robust simulators is essential for asking "what if?" questions and guiding policy in critical domains like healthcare and logistics. However, existing methods often struggle, either failing to generalize beyond historical data or, when using Large Language Models (LLMs), suffering from inaccuracies and poor empirical alignment. We introduce G-Sim, a hybrid framework that automates simulator construction by synergizing LLM-driven structural design with rigorous empirical calibration. G-Sim employs an LLM in an iterative loop to propose and refine a simulator's core components and causal relationships, guided by domain knowledge. This structure is then grounded in reality by estimating its parameters using flexible calibration techniques. Specifically, G-Sim can leverage methods that are both likelihood-free and gradient-free with respect to the simulator, such as gradient-free optimization for direct parameter estimation or simulation-based inference for obtaining a posterior distribution over parameters. This allows it to handle non-differentiable and stochastic simulators. By integrating domain priors with empirical evidence, G-Sim produces reliable, causally-informed simulators, mitigating data-inefficiency and enabling robust system-level interventions for complex decision-making.

cs.LG↗