Nawaf Alampara

Martiño Ríos-García, Nawaf Alampara, Chandan Gupta et al.

Large language model (LLM)-based systems are increasingly deployed to conduct scientific research autonomously, yet whether their reasoning adheres to the epistemic norms that make scientific inquiry self-correcting is poorly understood. Here, we evaluate LLM-based scientific agents across eight domains, spanning workflow execution to hypothesis-driven inquiry, through more than 25,000 agent runs and two complementary lenses: (i) a systematic performance analysis that decomposes the contributions of the base model and the agent scaffold, and (ii) a behavioral analysis of the epistemological structure of agent reasoning. We observe that the base model is the primary determinant of both performance and behavior, accounting for 41.4% of explained variance versus 1.5% for the scaffold. Across all configurations, evidence is ignored in 68% of traces, refutation-driven belief revision occurs in 26%, and convergent multi-test evidence is rare. The same reasoning pattern appears whether the agent executes a computational workflow or conducts hypothesis-driven inquiry. They persist even when agents receive near-complete successful reasoning trajectories as context, and the resulting unreliability compounds across repeated trials in epistemically demanding domains. Thus, current LLM-based agents execute scientific workflows but do not exhibit the epistemic patterns that characterize scientific reasoning. Outcome-based evaluation cannot detect these failures, and scaffold engineering alone cannot repair them. Until reasoning itself becomes a training target, the scientific knowledge produced by such agents cannot be justified by the process that generated it.

Martiño Ríos-García, Nawaf Alampara, Kevin Maik Jablonka

Counting should not depend on what is being counted; more generally, any algorithm's behavior should be invariant to the semantic content of its arguments. We introduce WhatCounts to test this property in isolation. Unlike prior work that conflates semantic sensitivity with reasoning complexity or prompt variation, WhatCounts is atomic: count items in an unambiguous, delimited list with no duplicates, distractors, or reasoning steps for different semantic types. Frontier LLMs show over 40% accuracy variation depending solely on what is being counted - cities versus chemicals, names versus symbols. Controlled ablations rule out confounds. The gap is semantic, and it shifts unpredictably with small amounts of unrelated fine-tuning. LLMs do not implement algorithms; they approximate them, and the approximation is argument-dependent. As we show with an agentic example, this has implications beyond counting: any LLM function may carry hidden dependencies on the meaning of its inputs.

Nawaf Alampara, Mara Schilling-Wilhelmi, Martiño Ríos-García et al.

Recent advancements in artificial intelligence have sparked interest in scientific assistants that could support researchers across the full spectrum of scientific workflows, from literature review to experimental design and data analysis. A key capability for such systems is the ability to process and reason about scientific information in both visual and textual forms - from interpreting spectroscopic data to understanding laboratory setups. Here, we introduce MaCBench, a comprehensive benchmark for evaluating how vision-language models handle real-world chemistry and materials science tasks across three core aspects: data extraction, experimental understanding, and results interpretation. Through a systematic evaluation of leading models, we find that while these systems show promising capabilities in basic perception tasks - achieving near-perfect performance in equipment identification and standardized data extraction - they exhibit fundamental limitations in spatial reasoning, cross-modal information synthesis, and multi-step logical inference. Our insights have important implications beyond chemistry and materials science, suggesting that developing reliable multimodal AI scientific assistants may require advances in curating suitable training data and approaches to training those models.

Adrian Mirza, Nawaf Alampara, Sreekanth Kunchapu et al.

Large language models (LLMs) have gained widespread interest due to their ability to process human language and perform tasks on which they have not been explicitly trained. However, we possess only a limited systematic understanding of the chemical capabilities of LLMs, which would be required to improve models and mitigate potential harm. Here, we introduce "ChemBench," an automated framework for evaluating the chemical knowledge and reasoning abilities of state-of-the-art LLMs against the expertise of chemists. We curated more than 2,700 question-answer pairs, evaluated leading open- and closed-source LLMs, and found that the best models outperformed the best human chemists in our study on average. However, the models struggle with some basic tasks and provide overconfident predictions. These findings reveal LLMs' impressive chemical capabilities while emphasizing the need for further research to improve their safety and usefulness. They also suggest adapting chemistry education and show the value of benchmarking frameworks for evaluating LLMs in specific domains.

Yoel Zimmermann, Adib Bazgir, Zartashia Afzal et al.

Here, we present the outcomes from the second Large Language Model (LLM) Hackathon for Applications in Materials Science and Chemistry, which engaged participants across global hybrid locations, resulting in 34 team submissions. The submissions spanned seven key application areas and demonstrated the diverse utility of LLMs for applications in (1) molecular and material property prediction; (2) molecular and material design; (3) automation and novel interfaces; (4) scientific communication and education; (5) research data management and automation; (6) hypothesis generation and evaluation; and (7) knowledge extraction and reasoning from scientific literature. Each team submission is presented in a summary table with links to the code and as brief papers in the appendix. Beyond team results, we discuss the hackathon event and its hybrid format, which included physical hubs in Toronto, Montreal, San Francisco, Berlin, Lausanne, and Tokyo, alongside a global online hub to enable local and virtual collaboration. Overall, the event highlighted significant improvements in LLM capabilities since the previous year's hackathon, suggesting continued expansion of LLMs for applications in materials science and chemistry research. These outcomes demonstrate the dual utility of LLMs as both multipurpose models for diverse machine learning tasks and platforms for rapid prototyping custom applications in scientific research.

Nawaf Alampara, Mara Schilling-Wilhelmi, Kevin Maik Jablonka

Measurements are fundamental to knowledge creation in science, enabling consistent sharing of findings and serving as the foundation for scientific discovery. As machine learning systems increasingly transform scientific fields, the question of how to effectively evaluate these systems becomes crucial for ensuring reliable progress. In this review, we examine the current state and future directions of evaluation frameworks for machine learning in science. We organize the review around a broadly applicable framework for evaluating machine learning systems through the lens of statistical measurement theory, using materials science as our primary context for examples and case studies. We identify key challenges common across machine learning evaluation such as construct validity, data quality issues, metric design limitations, and benchmark maintenance problems that can lead to phantom progress when evaluation frameworks fail to capture real-world performance needs. By examining both traditional benchmarks and emerging evaluation approaches, we demonstrate how evaluation choices fundamentally shape not only our measurements but also research priorities and scientific progress. These findings reveal the critical need for transparency in evaluation design and reporting, leading us to propose evaluation cards as a structured approach to documenting measurement choices and limitations. Our work highlights the importance of developing a more diverse toolbox of evaluation techniques for machine learning in materials science, while offering insights that can inform evaluation practices in other scientific domains where similar challenges exist.

Nawaf Alampara, Anagha Aneesh, Martiño Ríos-García et al.

Data-driven techniques have a large potential to transform and accelerate the chemical sciences. However, chemical sciences also pose the unique challenge of very diverse, small, fuzzy datasets that are difficult to leverage in conventional machine learning approaches completely. A new class of models, general-purpose models (GPMs) such as large language models, have shown the ability to solve tasks they have not been directly trained on, and to flexibly operate with low amounts of data in different formats. In this review, we discuss fundamental building principles of GPMs and review recent applications of those models in the chemical sciences across the entire scientific process. While many of these applications are still in the prototype phase, we expect that the increasing interest in GPMs will make many of them mature in the coming years.

Adrian Mirza, Nawaf Alampara, Martiño Ríos-García et al.

Foundation models have shown remarkable success across scientific domains, yet their impact in chemistry remains limited due to the absence of diverse, large-scale, high-quality datasets that reflect the field's multifaceted nature. We present the ChemPile, an open dataset containing over 75 billion tokens of curated chemical data, specifically built for training and evaluating general-purpose models in the chemical sciences. The dataset mirrors the human learning journey through chemistry -- from educational foundations to specialized expertise -- spanning multiple modalities and content types including structured data in diverse chemical representations (SMILES, SELFIES, IUPAC names, InChI, molecular renderings), scientific and educational text, executable code, and chemical images. ChemPile integrates foundational knowledge (textbooks, lecture notes), specialized expertise (scientific articles and language-interfaced data), visual understanding (molecular structures, diagrams), and advanced reasoning (problem-solving traces and code) -- mirroring how human chemists develop expertise through diverse learning materials and experiences. Constructed through hundreds of hours of expert curation, the ChemPile captures both foundational concepts and domain-specific complexity. We provide standardized training, validation, and test splits, enabling robust benchmarking. ChemPile is openly released via HuggingFace with a consistent API, permissive license, and detailed documentation. We hope the ChemPile will serve as a catalyst for chemical AI, enabling the development of the next generation of chemical foundation models.

Nawaf Alampara, Santiago Miret, Kevin Maik Jablonka

Predicting properties from coordinate-category data -- sets of vectors paired with categorical information -- is fundamental to computational science. In materials science, this challenge manifests as predicting properties like formation energies or elastic moduli from crystal structures comprising atomic positions (vectors) and element types (categorical information). While large language models (LLMs) have increasingly been applied to such tasks, with researchers encoding structural data as text, optimal strategies for achieving reliable predictions remain elusive. Here, we report fundamental limitations in LLM's ability to learn from coordinate information in coordinate-category data. Through systematic experiments using synthetic datasets with tunable coordinate and category contributions, combined with a comprehensive benchmarking framework (MatText) spanning multiple representations and model scales, we find that LLMs consistently fail to capture coordinate information while excelling at category patterns. This geometric blindness persists regardless of model size (up to 70B parameters), dataset scale (up to 2M structures), or text representation strategy. Our findings suggest immediate practical implications: for materials property prediction tasks dominated by structural effects, specialized geometric architectures consistently outperform LLMs by significant margins, as evidenced by a clear "GNN-LM wall" in performance benchmarks. Based on our analysis, we provide concrete guidelines for architecture selection in scientific machine learning, while highlighting the critical importance of understanding model inductive biases when tackling scientific prediction problems.