arXiv subjects

Samiah Hassan

Publications and source records attributed to Samiah Hassan.

3 recordsLinked to original sources

arXiv2026

Eutectic and peritectic equilibria in coherent binary alloys

This work extends the Cahn--Larch\'e thermodynamic framework to binary alloys in which two coherent solid phases coexist with an incoherent liquid and investigates how coherency strain energy modifies classical eutectic and peritectic equilibria. We derive equilibrium conditions for three-phase coexistence that include an elastic energy term dependent on the molar fractions of the solid phases and apply them to model binary eutectic and peritectic systems. We find that coherency stress transforms the eutectic point into a finite three-phase equilibrium field spanning a continuous range of compositions and temperatures. In contrast, coherency stress in peritectic systems progressively destabilizes the two-solid equilibrium without generating a stable three-phase field and can suppress the peritectic reaction entirely. This asymmetry is governed by the geometric relationship between the stress-free compositions of the phases: when the liquid composition lies between those of the two solids (eutectic configuration), the liquid serves as a thermodynamic buffer against the coherency penalty on the solid--solid pair; when it lies outside (peritectic configuration), no such mechanism is available. These results demonstrate that coherency stress can fundamentally alter three-phase equilibria involving a liquid and suggest that such effects may be significant in systems with large coherent misfits.

cond-mat.mtrl-sci↗

arXiv2025

CodeDistiller: Automatically Generating Code Libraries for Scientific Coding Agents

Automated Scientific Discovery (ASD) systems can help automatically generate and run code-based experiments, but their capabilities are limited by the code they can reliably generate from parametric knowledge alone. As a result, current systems either mutate a small number of manually-crafted experiment examples, or operate solely from parametric knowledge, limiting quality and reach. We introduce CodeDistiller, a system that automatically distills large collections of scientific Github repositories into a vetted library of working domain-specific code examples, allowing ASD agents to expand their capabilities without manual effort. Using a combination of automatic and domain-expert evaluation on 250 materials science repositories, we find the best model is capable of producing functional examples for 74% of repositories, while our downstream evaluation shows an ASD agent augmented with a CodeDistiller generated library produces more accurate, complete, and scientifically sound experiments than an agent with only general materials-science code examples. We also evaluate LLM-as-a-judge ratings against domain-expert ratings in an A/B testing paradigm, finding moderate agreement and suggesting that inexpensive proxy metrics may be feasible for evaluating scientific discovery systems at scale.

cs.AI↗

arXiv2025

Matter-of-Fact: A Benchmark for Verifying the Feasibility of Literature-Supported Claims in Materials Science

Contemporary approaches to assisted scientific discovery use language models to automatically generate large numbers of potential hypothesis to test, while also automatically generating code-based experiments to test those hypotheses. While hypotheses can be comparatively inexpensive to generate, automated experiments can be costly, particularly when run at scale (i.e. thousands of experiments). Developing the capacity to filter hypotheses based on their feasibility would allow discovery systems to run at scale, while increasing their likelihood of making significant discoveries. In this work we introduce Matter-of-Fact, a challenge dataset for determining the feasibility of hypotheses framed as claims, while operationalizing feasibility assessment as a temporally-filtered claim verification task using backtesting. Matter-of-Fact includes 8.4k claims extracted from scientific articles spanning four high-impact contemporary materials science topics, including superconductors, semiconductors, batteries, and aerospace materials, while including qualitative and quantitative claims from theoretical, experimental, and code/simulation results. We show that strong baselines that include retrieval augmented generation over scientific literature and code generation fail to exceed 72% performance on this task (chance performance is 50%), while domain-expert verification suggests nearly all are solvable -- highlighting both the difficulty of this task for current models, and the potential to accelerate scientific discovery by making near-term progress.

cs.AI↗