Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Yang, Jian, Zhang, Wei, Wu, Jiajun, Cheng, Junhang, Zheng, Tuney, Xu, Fanglin, Gu, Weicheng, Jing, Lin, Du, Yaxin, Li, Joseph, Li, Yizhi, Xing, Yan, Hao, Chuan, Tao, Ran, Gong, Ruihao, Liu, Aishan, Li, Zhoujun, Tang, Mingjie, Lin, Chenghua, Chen, Siheng, Zhao, Wayne Xin, Liu, Xianglong, Zhou, Ming, Dai, Bryan, Lv, Weifeng
Format: Preprint
Veröffentlicht: 2026
Schlagworte:
Online-Zugang:https://arxiv.org/abs/2604.03144
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866913002168516608
author Yang, Jian
Zhang, Wei
Wu, Jiajun
Cheng, Junhang
Zheng, Tuney
Xu, Fanglin
Gu, Weicheng
Jing, Lin
Du, Yaxin
Li, Joseph
Li, Yizhi
Xing, Yan
Hao, Chuan
Tao, Ran
Gong, Ruihao
Liu, Aishan
Li, Zhoujun
Tang, Mingjie
Lin, Chenghua
Chen, Siheng
Zhao, Wayne Xin
Liu, Xianglong
Zhou, Ming
Dai, Bryan
Lv, Weifeng
author_facet Yang, Jian
Zhang, Wei
Wu, Jiajun
Cheng, Junhang
Zheng, Tuney
Xu, Fanglin
Gu, Weicheng
Jing, Lin
Du, Yaxin
Li, Joseph
Li, Yizhi
Xing, Yan
Hao, Chuan
Tao, Ran
Gong, Ruihao
Liu, Aishan
Li, Zhoujun
Tang, Mingjie
Lin, Chenghua
Chen, Siheng
Zhao, Wayne Xin
Liu, Xianglong
Zhou, Ming
Dai, Bryan
Lv, Weifeng
contents 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
format Preprint
id arxiv_https___arxiv_org_abs_2604_03144
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle InCoder-32B-Thinking: Industrial Code World Model for Thinking
Yang, Jian
Zhang, Wei
Wu, Jiajun
Cheng, Junhang
Zheng, Tuney
Xu, Fanglin
Gu, Weicheng
Jing, Lin
Du, Yaxin
Li, Joseph
Li, Yizhi
Xing, Yan
Hao, Chuan
Tao, Ran
Gong, Ruihao
Liu, Aishan
Li, Zhoujun
Tang, Mingjie
Lin, Chenghua
Chen, Siheng
Zhao, Wayne Xin
Liu, Xianglong
Zhou, Ming
Dai, Bryan
Lv, Weifeng
Hardware Architecture
Artificial Intelligence
Computation and Language
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
title InCoder-32B-Thinking: Industrial Code World Model for Thinking
topic Hardware Architecture
Artificial Intelligence
Computation and Language
url https://arxiv.org/abs/2604.03144