Write Once, Run Everywhere: The Axon DSL for Shape-Safe and Framework-Agnostic LLM Architectures

• URL: http://arxiv.org/abs/2608.19889v1
• Date: Thu, 20 Aug 2026 10:56:34 GMT
• Title: Write Once, Run Everywhere: The Axon DSL for Shape-Safe and Framework-Agnostic LLM Architectures
• Abstract summary: Axon is a strongly typed domain-specific language with Haskell-like syntax.<n>By basing collaboration on a language specification rather than a specific framework's vision, Axon fosters open cooperation.<n>In 467 benchmarking experiments on models ranging from 135M to 32B parameters, we demonstrate median speedups of 7% on PyTorch, 12% on PyTorch with Triton, 91% on JAX, and 107% on MLX.
• Score: 4.307697295291473
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: The entire ecosystem of open-source language models effectively relies on a single platform. What if this platform was forced to shut down tomorrow? Implementing and maintaining efficient model definitions and translating them between different training and inference regimes is a resource-heavy task that severely limits model efficiency and portability, hindering both scaling and deployment. Here, we present Axon, a strongly typed domain-specific language with Haskell-like syntax, that enables a write-once, run everywhere paradigm for LLM architectures. By basing collaboration on a language specification rather than a specific framework's vision, Axon fosters open cooperation and empowers researchers to implement highly specialized architectures without giving up optimization infrastructure or accepting deployment lock-in. Axon allows for concise, auditable specifications that can be automatically compiled to standalone implementations for leading frameworks: PyTorch, PyTorch with Triton, JAX, MLX and vLLM. In 467 inference benchmarking experiments on models ranging from 135M to 32B parameters, we demonstrate median speedups of 7% on PyTorch, 12% on PyTorch with Triton, 91% on JAX, and 107% on MLX, compared to the reference implementations from Transformers. When deployed as native vLLM architectures with PagedAttention and KV-cache, Axon models achieve a 58% median speedup over Transformers implementations.

Related papers

• ClawArena-Team: Benchmarking Subagent Orchestration and Dynamic Workflows in Language-Model Agents [61.184954205350984]
  ClawArena-Team is a benchmark of 41 multi-turn, multimodal, multi-agent scenarios spanning 258 evaluation rounds and 72 staged updates.<n>All scoring is execution-based with no LLM judge: an overall score -- the Subagent-Management Score (SMS) -- multiplies task correctness by a least-privilege and modality-routing factor.
  arXiv  Detail & Related papers  (2026-06-30T06:06:08Z)
• Agentic Framework for Deep Learning workload migration via In-Context Learning [0.5496900364175868]
  Large Language Models (LLMs) are prone to mistakes for exacting operations.<n>We propose a fully autonomous system that combines In-Context Learning (ICL) with oracle-driven self-ging.<n>Our pipeline achieves 91% numerical equivalence (compared to baseline: 9%, instruction + self-ging: 27%) on neural modules.
  arXiv  Detail & Related papers  (2026-06-14T19:41:57Z)
• OneDrive: Unified Multi-Paradigm Driving with Vision-Language-Action Models [69.2503510410147]
  We present a unified autonomous driving framework built upon a pretrained VLM.<n>We demonstrate that pretrained VLM attention exhibits strong transferability beyond pure language modeling.<n>Experiments on end-to-end autonomous driving benchmarks demonstrate state-of-the-art performance.
  arXiv  Detail & Related papers  (2026-04-20T07:50:00Z)
• From Translation to Superset: Benchmark-Driven Evolution of a Production AI Agent from Rust to Python [2.7324157162184157]
  We present a methodology for LLM-assisted continuous code translation in which a large language model translates a production Rust objective into Python.<n>We demonstrate that the Python port resolves 59/80 SWE-bench Verified tasks (73.8%) versus Rust's 56/80 (70.0%), confirming near-parity on real-world agentic tasks.<n>Our evaluation shows that for LLM-based agents where API latency dominates, Python's expressiveness yields a 15.9x code reduction with negligible performance cost.
  arXiv  Detail & Related papers  (2026-04-13T14:21:44Z)
• SAGAI-MID: A Generative AI-Driven Middleware for Dynamic Runtime Interoperability [1.2744523252873352]
  SAGAI-MID is a FastAPI that uses large language models (LLMs) to dynamically detect and resolve schema mismatches at runtime.<n>We evaluate SAGAI-MID on 10 interoperability scenarios spanning version migration, IoT-to-analytics bridging, and protocol conversion across six LLMs from two providers.
  arXiv  Detail & Related papers  (2026-03-30T17:46:41Z)
• Interfaze: The Future of AI is built on Task-Specific Small Models [0.25489046505746704]
  We present Interfaze, a system that treats modern LLM applications as a problem of building and acting over context.<n>A thin controller sits on top of this stack and exposes a single, OpenAI-style endpoint.<n>On this architecture, Interfaze-Beta achieves 83.6% on MMLU-Pro, 91.4% on MMLU, 81.3% on GPQA-Diamond, 57.8% on LiveCodeBench v5, and 90.0% on AIME-2025.
  arXiv  Detail & Related papers  (2026-02-04T00:36:37Z)
• Spark-LLM-Eval: A Distributed Framework for Statistically Rigorous Large Language Model Evaluation [0.0]
  We present Spark-LLM-Eval, a distributed evaluation framework built on Apache Spark.<n>The framework and all evaluation code are available as open source.
  arXiv  Detail & Related papers  (2026-01-18T04:34:39Z)
• A Multi-Language Object-Oriented Programming Benchmark for Large Language Models [61.267115598083315]
  A survey of 35 existing benchmarks uncovers three major imbalances.<n>85.7% focus on a single programming language.<n>94.3% target only function-level or statement-level tasks.<n>Over 80% include fewer than ten test cases on average.
  arXiv  Detail & Related papers  (2025-09-30T11:30:08Z)
• SwingArena: Competitive Programming Arena for Long-context GitHub Issue Solving [90.32201622392137]
  We present SwingArena, a competitive evaluation framework for Large Language Models (LLMs)<n>Unlike traditional static benchmarks, SwingArena models the collaborative process of software by pairing LLMs as iterations, who generate patches, and reviewers, who create test cases and verify the patches through continuous integration (CI) pipelines.
  arXiv  Detail & Related papers  (2025-05-29T18:28:02Z)
• CRAB: Cross-environment Agent Benchmark for Multimodal Language Model Agents [49.68117560675367]
  Crab is the first benchmark framework designed to support cross-environment tasks.<n>Our framework supports multiple devices and can be easily extended to any environment with a Python interface.<n>The experimental results demonstrate that the single agent with GPT-4o achieves the best completion ratio of 38.01%.
  arXiv  Detail & Related papers  (2024-07-01T17:55:04Z)
• Cheaply Evaluating Inference Efficiency Metrics for Autoregressive Transformer APIs [66.30706841821123]
  Large language models (LLMs) power many state-of-the-art systems in natural language processing. LLMs are extremely computationally expensive, even at inference time. We propose a new metric for comparing inference efficiency across models.
  arXiv  Detail & Related papers  (2023-05-03T21:51:42Z)

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.