cs.AIcond-mat.mtrl-sciphysics.comp-phEmpirical
Grounded autonomous research: a fault-tolerant LLM pipeline from corpus to manuscript in frontier computational physics
Jul 2, 2026
An autonomous research agent is developed to automate end-to-end LLM in high-stakes scientific domains, specifically condensed matter physics, by mapping the corpus, calibrating methodology, conducting computations, and writing a publication-grade manuscript.
Develops an autonomous research agent that can read scientific papers, come up with new research directions, conduct computations, and write publication-grade manuscripts in high-stakes scientific domains.
Keywords
Before reading this…
Applications
To understand this paper, make sure you know these concepts first:
Autonomous-research agents have demonstrated end-to-end LLM automation in machine-learning sandboxes where execution provides calibration. Frontier physical science differs categorically: physical reasoning underlies every methodology choice, toolchains are often underdocumented, and calibration must come from external literature anchors - which unscaffolded agents cite but do not confront, hallucinating plausible, unverifiable results from internal priors. We present a pipeline that runs end-to-end from a corpus of 11,083 recent condensed-matter physics arXiv papers to a publication-grade manuscript with three substantive physics findings (here on altermagnetic piezomagnetism): the agent autonomously conceives a research direction by mapping the corpus, calibrates methodology by reproducing published references, conducts novel first-principles computations, and writes the manuscript - grounded in literature throughout, across 47 fresh-context sessions in six phases sharing only on-disk state, with 2,162 literature-consultation events. Fault tolerance emerges from redundancy: fresh-context isolation, distributed grounding, and adversarial review catch what any single session misses; pre- and post-pilot stages are fully autonomous, and pilot requires bounded human intervention only at reproduction failures - operational knowledge curation, not scientific direction. Two paired failure modes - a pre-architecture baseline and a no-pilot ablation - isolate structurally enforced numerical confrontation at calibration checkpoints as the operative grounding mechanism. The primitives, characterized failure modes, and quantified intervention pattern lay a foundation for autonomous research in high-stakes scientific domains beyond computational physics.