Compile by Training: Turning Natural-Language Specifications into Local Neural Functions

• URL: http://arxiv.org/abs/2609.04199v1
• Date: Thu, 03 Sep 2026 17:59:49 GMT
• Title: Compile by Training: Turning Natural-Language Specifications into Local Neural Functions
• Abstract summary: compile by training turns a natural-language specification into a reusable neural function.<n>On FuzzyBench-Hard, a subset on which the Program-as-Weights fast compiler produced no exact matches, compile by training reaches 83.6% semantic accuracy.
• Score: 8.34539885321864
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Many recurring text functions are easy to describe but difficult to implement with rules, while calling a large remote model for every input introduces repeated cost, latency, and dependency on a provider. We present compile by training, which turns a natural-language specification into a reusable neural function. At compile time, teacher models generate task-specific examples that are used to train a small adapter for a compact interpreter. The resulting function runs without the teachers and can be stored, versioned, and composed like ordinary software. On FuzzyBench-Hard, a subset on which the Program-as-Weights fast compiler produced no exact matches, compile by training reaches 83.6% semantic accuracy. This higher accuracy comes with a higher compile-time cost: roughly a minute rather than seconds for the fast compiler. We deploy the compiler in a public interactive service and demonstrate compiled functions in a multi-site website helper, a language-controlled 3D avatar, and a bidirectional English-Claudish translator.

Related papers

• Neural Code Translation of Legacy Code: APL to C# [0.0]
  This paper investigates the translation of APL into C# using large language models.<n>We introduce a novel framework for APL-to-C# translation by comparing three guided strategies.<n>Our results demonstrate that neural code translation can successfully bridge the gap between APL and C# for a wide range of programs.
  arXiv  Detail & Related papers  (2026-05-12T12:11:33Z)
• Decaf: Improving Neural Decompilation with Automatic Feedback and Search [64.02615739798277]
  High-level syntax, identifiers, and custom data types are generally lost as the compiler translates human-readable code to low-level machine code.<n>Generative AI models are a natural fit for reconstructing high-level syntax, identifiers, and types.<n>We argue that compiler feedback can be used to dramatically improve the semantic correctness of neural decompiler outputs via search.
  arXiv  Detail & Related papers  (2026-05-12T04:21:26Z)
• Small Language Models as Compiler Experts: Auto-Parallelization for Heterogeneous Systems [0.0]
  This paper presents a comprehensive evaluation of small (approximately 1B parameter) language-model-driven compiler auto-parallelization.<n>We evaluate three models: gemma3, llama3.2, and qwen2.5, using six reasoning strategies across 11 real-world kernels.<n>Across 376 total evaluations, the proposed approach achieves an average speedup of 6.81x and a peak performance of 43.25x on convolution operations.
  arXiv  Detail & Related papers  (2025-12-22T10:34:45Z)
• Guided Tensor Lifting [54.10411390218929]
  Domain-specific languages (s) for machine learning are revolutionizing the speed and efficiency of machine learning workloads.<n>To take advantage of these capabilities, a user must first translate their legacy code from the language it is currently written in, into the new DSL.<n>Process of automatically lifting code into these DSLs has been identified by several recent works, which propose program synthesis as a solution.
  arXiv  Detail & Related papers  (2025-04-28T12:00:10Z)
• Developing a Modular Compiler for a Subset of a C-like Language [0.0]
  The paper introduces the development of a modular compiler for a subset of a C-like language.<n>This modular approach will allow developers to modify a language by adding or removing subsets as required, resulting in a minimal and memory-efficient compiler.
  arXiv  Detail & Related papers  (2025-01-08T13:42:54Z)
• Dcc --help: Generating Context-Aware Compiler Error Explanations with Large Language Models [53.04357141450459]
  dcc --help was deployed to our CS1 and CS2 courses, with 2,565 students using the tool over 64,000 times in ten weeks. We found that the LLM-generated explanations were conceptually accurate in 90% of compile-time and 75% of run-time cases, but often disregarded the instruction not to provide solutions in code.
  arXiv  Detail & Related papers  (2023-08-23T02:36:19Z)
• Decomposed Prompting for Machine Translation Between Related Languages using Large Language Models [55.35106713257871]
  We introduce DecoMT, a novel approach of few-shot prompting that decomposes the translation process into a sequence of word chunk translations. We show that DecoMT outperforms the strong few-shot prompting BLOOM model with an average improvement of 8 chrF++ scores across the examined languages.
  arXiv  Detail & Related papers  (2023-05-22T14:52:47Z)
• QParallel: Explicit Parallelism for Programming Quantum Computers [62.10004571940546]
  We present a language extension for parallel quantum programming. QParallel removes ambiguities concerning parallelism in current quantum programming languages. We introduce a tool that guides programmers in the placement of parallel regions by identifying the subroutines that profit most from parallelization.
  arXiv  Detail & Related papers  (2022-10-07T16:35:16Z)
• Compilable Neural Code Generation with Compiler Feedback [43.97362484564799]
  This paper proposes a three-stage pipeline for compilable code generation, including language model fine-tuning, compilability reinforcement, and compilability discrimination. Experiments on two code generation tasks demonstrate the effectiveness of our proposed approach, improving the success rate of compilation from 44.18 to 89.18 on average and from 70.3 to 96.2 in text-to-code generation, respectively.
  arXiv  Detail & Related papers  (2022-03-10T03:15:17Z)
• AVATAR: A Parallel Corpus for Java-Python Program Translation [77.86173793901139]
  Program translation refers to migrating source code from one language to another. We present AVATAR, a collection of 9,515 programming problems and their solutions written in two popular languages, Java and Python.
  arXiv  Detail & Related papers  (2021-08-26T05:44:20Z)
• Leveraging Language to Learn Program Abstractions and Search Heuristics [66.28391181268645]
  We introduce LAPS (Language for Abstraction and Program Search), a technique for using natural language annotations to guide joint learning of libraries and neurally-guided search models for synthesis. When integrated into a state-of-the-art library learning system (DreamCoder), LAPS produces higher-quality libraries and improves search efficiency and generalization.
  arXiv  Detail & Related papers  (2021-06-18T15:08:47Z)
• LazyTensor: combining eager execution with domain-specific compilers [0.0]
  LazyTensor is a technique to target domain compilers without sacrificing define-by-run ergonomics. Initially developed to support PyTorch on Cloud TPUs, along with a substantially shared implementation technique, has been used by Swift for specific accelerators.
  arXiv  Detail & Related papers  (2021-02-26T02:22:16Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.