Papers where the original authors have engaged with our plain-language explanation.
For every paper on this page, at least one of the original authors has seen our plain-language explanation and engaged with it — either confirming it reads accurately or requesting corrections that we then applied. An endorsement does not mean the authors formally approve every sentence, but it does mean the explanation has passed the eyes of the people who wrote the paper.
438 papers reviewed by authors · 411–420 / 438
This paper introduces a post-training paradigm that leverages knowledge graphs as implicit reward models to guide large language models in learning compositional reasoning from axiomatic facts, enabling a 14B model to outperform frontier systems on complex multi-hop medical queries through path-derived supervision.
This paper demonstrates that directional CDNV (decision-axis variance) is the key geometric quantity governing both strong few-shot transfer and low multitask interference in self-supervised learning, as its minimization during pretraining leads to tight generalization bounds and nearly orthogonal decision axes across tasks.
Contrary to the conventional expectation that electrowetting promotes droplet spreading, this study reveals that applying DC voltage to specific microtextured, lubricant-infused surfaces in a Cassie wetting state induces rapid tangential droplet ejection due to unbalanced electrocapillary forces and minimal pinning, offering a new paradigm for controlled droplet transport.
This study demonstrates that a Pechini sol-gel synthesis method with cationic molecular mixing can efficiently produce high-quality Li-doped Bi-2223 superconductors, achieving a maximum critical temperature of 111.4 K at 5 mol% Li doping while revealing unique layer-by-layer crystalline growth and flux creep mechanisms.
This paper reviews how recent advancements in Large Momentum Effective Theory (LaMET), including improved renormalization schemes, higher-order matching kernels, and novel lattice operators, are enhancing the precision and reliability of first-principles lattice QCD calculations for proton parton distributions.
This paper predicts that upcoming Gaia data releases will detect tens to hundreds of circumbinary planets, revealing a yield significantly lower than pre-launch estimates but highly sensitive to the underlying mass and orbital period distributions, while also offering the potential to validate or refute candidates around post-common-envelope binaries.
This paper introduces the Eclipse-Yarkovsky effect, a thermal torque arising from asymmetric particle emission during planetary eclipses, and demonstrates through a continuum framework that it drives outward ring transport and can form sharp inner edges, while a competing planetary-Yarkovsky effect may cause inward migration in close-in rings.
This paper proposes that a predicted fourth generation of fermions, characterized by unique electric charges and conserved quantum numbers that prevent interaction with known matter, forms neutral composites that serve as candidates for both super-massive galactic core dark matter and galactic halo dark matter.
Odin is the first production-deployed autonomous graph intelligence engine that utilizes the novel COMPASS multi-signal scoring metric to guide efficient, traceable, and hallucination-free pattern discovery in knowledge graphs, successfully addressing local community entrapment in regulated industries like healthcare and insurance.
Addressing the breakdown of Bloch's theorem in spatially varying composites, this paper establishes a foundational ab initio quantum theoretical framework for functionally graded materials that derives effective field equations for modulated Bloch states, revealing non-tensorial electromagnetic properties and enabling the predictive design of optimized electronic devices such as graded p-n junctions.