The paper introduces "compile by training," a method that converts natural-language specifications into reusable neural functions. By generating task-specific examples with teacher models at compile time, a small adapter is trained for a compact interpreter, eliminating the need for remote model calls during execution. The approach achieves 83.6% semantic accuracy on FuzzyBench-Hard, though it incurs a higher compile-time cost of about a minute, and is demonstrated in applications such as a multi-site website helper, a language-controlled 3D avatar, and a bidirectional English‑Claudish translator.
By Yuntian Deng, Pengyu Nie, Stuart Shieber
arXiv:2608. 19889v1 Announce Type: new Abstract: The entire ecosystem of open-source language models effectively relies on a single platform.
By Jacob Nielsen, Danial Namazifard, Lukas Galke Poech, Peter Schneider-Kamp
arXiv:2606. 10087v1 Announce Type: cross Abstract: Pre-training on raw code teaches syntax but provides sparse signal for diverse real-world task formats.
By Ankit Gupta, Aditya Prasad, Rameswar Panda
arXiv:2607. 13921v1 Announce Type: cross Abstract: Languages with rich static semantics, such as Rust, provide stronger guarantees for AI-generated code, but their strictness makes generation more difficult.
By Niels M\"undler-Sasahara, Hristo Venev, Dawn Song, Martin Vechev, Jingxuan He
arXiv:2607. 03574v1 Announce Type: cross Abstract: AI systems increasingly propose executable scientific models whose value depends on both their symbolic structure and their fitted continuous parameters.
By Lucas Sheneman
arXiv:2607. 22595v1 Announce Type: new Abstract: Mechanistic interpretability (MI) has emerged as a powerful approach for analyzing and intervening in inference computations, with a growing number of applications such as jailbreak attempt detection, truthfulness evaluation, and hallucination detection.
By Michael Blum, Mark Silberstein, Yaniv David