Read original ↗
paperarXivTrust 82 · PrimaryPublished 1mo agoLive · 1mo ago

Grounded autonomous research: a fault-tolerant LLM pipeline from corpus to manuscript in frontier computational physics

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 man

Lineage graph

Paper → model → repo connections mined from source citations (Tier-1 exact match).

Why these links exist

Every edge carries a method, confidence, and the source snippet that justified it — so bad links are debuggable.

Implements

repochrisliu298/awesome-llm-unlearningreposcikit-learn/scikit-learn

Covers

newsDeepSeek open sources DSpark, a new framework to speed up LLM inference by up to 85%newsIEEE Rolls Out Large Language Models Virtual Training CoursenewsHypothetically speaking...

authored (incoming)

personHaonan Huang

Implements (incoming)

repojanosh/matbench-discoveryrepoarogozhnikov/hep_ml

Covers (incoming)

newsUniFFBench: evaluating universal machine learning force fields against experimental measurements - NaturenewsAI-Assisted Scientific Writing: How to Use LLMs Without Compromising Research Integrity - Technology NetworksnewsNeed Arxiv endorsement for CS.LG to publish a preprint about Learning Stable Latent Manifolds from noisy sensor data /Observation space [P]

Related across the graph

repochrisliu298/awesome-llm-unlearningnewsDeepSeek open sources DSpark, a new framework to speed up LLM inference by up to 85%personHaonan Huangreposcikit-learn/scikit-learnrepoarogozhnikov/hep_mlrepojanosh/matbench-discoverynewsUniFFBench: evaluating universal machine learning force fields against experimental measurements - NaturenewsAI-Assisted Scientific Writing: How to Use LLMs Without Compromising Research Integrity - Technology NetworksnewsHypothetically speaking...newsIEEE Rolls Out Large Language Models Virtual Training CoursenewsNeed Arxiv endorsement for CS.LG to publish a preprint about Learning Stable Latent Manifolds from noisy sensor data /Observation space [P]

Topics

cs.AI