让代码模型学会像工程师一样思考硬件约束。
InCoder-32B-Thinking: Industrial Code World Model for Thinking
- 基于错误驱动的思维链生成工业级代码推理轨迹
- 在14个通用与9个工业基准上均达顶尖开源性能
- 适合芯片、嵌入式系统等硬件敏感场景开发者使用
芯片设计、GPU优化和嵌入式系统等工业软件开发缺乏展示工程师如何思考硬件约束与时序语义的推理过程。本文提出 InCoder-32B-Thinking，通过在错误驱动的思维链（ECoT）合成框架下训练，结合工业代码世界模型（ICWM）生成推理轨迹。ECoT通过多轮对话与环境错误反馈合成思维链，显式建模纠错过程；ICWM基于Verilog仿真、GPU性能分析等领域的执行轨迹训练，学习代码对硬件行为的影响机制，并能在编译前预测执行结果实现自验证。所有合成推理轨迹经领域工具链验证，数据分布符合工业任务真实推理深度。在14个通用基准（LiveCodeBench v5，81.3%）与9个工业基准（CAD-Coder，84.0%；KernelBench，38.0%）上，InCoder-32B-Thinking 均达到顶级开源水平。
原文摘要 · Abstract (English)
Industrial software development across chip design, GPU optimization, and embedded systems lacks expert reasoning traces showing how engineers reason about hardware constraints and timing semantics. In this work, we propose InCoder-32B-Thinking, trained on the data from the Error-driven Chain-of-Thought (ECoT) synthesis framework with an industrial code world model (ICWM) to generate reasoning traces. Specifically, ECoT generates reasoning chains by synthesizing the thinking content from multi-turn dialogue with environmental error feedback, explicitly modeling the error-correction process. ICWM is trained on domain-specific execution traces from Verilog simulation, GPU profiling, etc., learns the causal dynamics of how code affects hardware behavior, and enables self-verification by predicting execution outcomes before actual compilation. All synthesized reasoning traces are validated through domain toolchains, creating training data matching the natural reasoning depth distribution of industrial tasks. Evaluation on 14 general (81.3% on LiveCodeBench v5) and 9 industrial benchmarks (84.0% in CAD-Coder and 38.0% on KernelBench) shows InCoder-32B-Thinking achieves top-tier open-source results across all domains.GPU Optimization
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。