Numbat: Building and Verifying a Self-Contained Machine-Learning Stack

• URL: http://arxiv.org/abs/2609.10632v1
• Date: Wed, 09 Sep 2026 08:36:12 GMT
• Title: Numbat: Building and Verifying a Self-Contained Machine-Learning Stack
• Abstract summary: numbat is a machine-learning stack written in one general-purpose language (Zig) with no third-party dependencies.<n>We report on the construction and verification of numbat.
• Score: 0.0
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Machine-learning systems are built almost exclusively on a few large Python-orchestrated frameworks, and they inherit those stacks' engineering costs: environments of hundreds of version-coupled packages, separate export toolchains for deployment, and the split between the language research is written in and the language products ship in. We report on the construction and verification of numbat, a machine-learning stack written in one general-purpose language (Zig) with no third-party runtime dependencies. The stack spans tensor computation, automatic differentiation, neural-network modules, mixed precision, multi-GPU training, data loading and monitoring; an SDK exposes it behind a stable, additively versioned C ABI of over 1,400 entry points, with bindings for six languages; and its clinical domain planes encode regulatory requirements as executable acceptance gates rather than documentation. Verifying such a stack is the harder half of building it: a defective training run rarely fails, it converges quietly to a slightly worse model. We treat a widely used reference implementation as an executable specification and verify against it at five levels, from operator gradient checks to an automated trajectory gate against a same-machine reference run - the arrangement our companion study formalizes as a trajectory-level differential oracle. The protocol surfaced ten silent recipe divergences, which we catalog with mechanisms and symptoms. As the acceptance test, we train a 25.9M-parameter detector of the YOLOv8m class from random initialization on COCO 2017 for the full 500-epoch schedule: the exported weights score 0.4956 mAP50-95 under the official protocol, scored by the reference stack's own validator (published endpoint 0.502), with single-GPU step time at parity on identical hardware. Weights, per-epoch metrics and the full run manifest are released.

Related papers

• Cross-Stack Validation of Language-Model Training: A Clinical Fine-Tuning Case Study [0.0]
  We study whether independently implemented training stacks can serve as differential oracles for a whole fine-tuning pipeline.<n>We apply it to a LoRA adaptation of Qwen3-0.6B over 168,574 clinical question-answer pairs.
  arXiv  Detail & Related papers  (2026-08-25T08:55:45Z)
• MergeSE: Post-Hoc Model Merging for Software Engineering Tasks Without Retraining [9.757638656919832]
  We present textbfMergeSE, an open-source CLI and web tool for training-free model merging of HFace encoder checkpoints.<n> MergeSE supports SE classification more broadly through a built-in registry of nine task types.<n>It supports five merging algorithms, including TIES, DARE-TIES, Wudi, PCB, and averaging.
  arXiv  Detail & Related papers  (2026-08-04T19:39:02Z)
• SCDBench: A Benchmark for LLM-Based Smart Contract Decompilers [55.39407031861402]
  We introduce SCDBench, a dataset and benchmark methodology for smart contract decompilation.<n>The dataset contains 600 real-world Solidity contracts with paired bytecode inputs, ground-truth source code, and replayable semantic checkpoints.<n>We evaluate Claude Opus 4.7, GPT-5.3-Codex, and GLM-5 in a zero-shot decompilation setting, including GLM-5 variants with and without extended reasoning.
  arXiv  Detail & Related papers  (2026-05-27T20:08:47Z)
• Context-Instrumental Data Distillation for Kubernetes Manifest Generation: Method and Experimental Evaluation [36.94429692322632]
  This paper examines the specialization of Small Language Models with up to 4 billion parameters for generating artifacts in domain-specific languages (AML)<n>We propose the context-instrumental data distillation method: the source corpus is formed through synthetic generation and, in an extended scheme, through reverse instruction generation from real Y files.
  arXiv  Detail & Related papers  (2026-05-25T13:30:38Z)
• ChipMATE: Multi-Agent Training via Reinforcement Learning for Enhanced RTL Generation [55.947962672433675]
  ChipMATE is the first self-trained multi-agent framework for RTL generation.<n>Inspired by industrial practice, ChipMATE pairs a Verilog agent with a Python reference-model agent that mutually verify each other's outputs.<n>ChipMATE achieves 75.0% and 80.1% pass@1 on VerilogEval V2 with 4B and 9B base models.
  arXiv  Detail & Related papers  (2026-05-13T01:04:21Z)
• 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)
• LIDARLearn: A Unified Deep Learning Library for 3D Point Cloud Classification, Segmentation, and Self-Supervised Representation Learning [1.2599533416395765]
  Three-dimensional (3D) point cloud analysis has become central to applications ranging from autonomous driving and robotics to forestry and ecological monitoring.<n>lib is a unified, PyTorch library that integrates over 55 model configurations covering 29 supervised architectures, seven SSL pre-training methods, and five PEFT strategies.<n>lib provides standardised training runners, cross-validation with stratified $K$-fold splitting, automated/CSV table generation, built-in Friedman/Nemenyi statistical testing with critical-difference diagrams for rigorous multi-model comparison, and a comprehensive test suite with 2,200+ automated tests
  arXiv  Detail & Related papers  (2026-04-12T19:10:12Z)
• VibeTensor: System Software for Deep Learning, Fully Generated by AI Agents [42.56489784841984]
  "fully generated" refers to code provenance: implementation changes were produced and applied as agent-proposed diffs.<n>We describe the architecture, summarize the workflow used to produce and validate the system, and evaluate the artifact.
  arXiv  Detail & Related papers  (2026-01-21T19:29:00Z)
• Commit0: Library Generation from Scratch [77.38414688148006]
  Commit0 is a benchmark that challenges AI agents to write libraries from scratch.<n>Agents are provided with a specification document outlining the library's API as well as a suite of interactive unit tests.<n> Commit0 also offers an interactive environment where models receive static analysis and execution feedback on the code they generate.
  arXiv  Detail & Related papers  (2024-12-02T18:11:30Z)
• 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)
• QAmeleon: Multilingual QA with Only 5 Examples [71.80611036543633]
  We show how to leverage pre-trained language models under a few-shot learning setting. Our approach, QAmeleon, uses a PLM to automatically generate multilingual data upon which QA models are trained. Prompt tuning the PLM for data synthesis with only five examples per language delivers accuracy superior to translation-based baselines.
  arXiv  Detail & Related papers  (2022-11-15T16:14:39Z)

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.