Regression Language Models for Code

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Akhauri, Yash, Song, Xingyou, Wongpanich, Arissa, Lewandowski, Bryan, Abdelfattah, Mohamed S.
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866917539297099776
author Akhauri, Yash
Song, Xingyou
Wongpanich, Arissa
Lewandowski, Bryan
Abdelfattah, Mohamed S.
author_facet Akhauri, Yash
Song, Xingyou
Wongpanich, Arissa
Lewandowski, Bryan
Abdelfattah, Mohamed S.
contents We study code-to-metric regression: predicting numeric outcomes of code executions, a challenging task due to the open-ended nature of programming languages. While prior methods have resorted to heavy and domain-specific feature engineering, we show that a single unified Regression Language Model (RLM) using a frozen LLM encoder can simultaneously predict directly from text, (i) the memory footprint of code across multiple high-level languages such as Python and C++, (ii) the latency of Triton GPU kernels, and (iii) the accuracy and speed of trained neural networks represented in ONNX. In particular, a relatively small 300M parameter RLM based on T5Gemma, obtains $>$0.9 Spearman-rank on competitive programming submissions from APPS, and a single unified model achieves $>$0.5 average Spearman-rank across 17 separate languages from CodeNet. Furthermore, the RLM can obtain the highest average Kendall-Tau of 0.46 on five classic NAS design spaces previously dominated by graph neural networks, and simultaneously predict architecture latencies on numerous hardware platforms.
format Preprint
id arxiv_https___arxiv_org_abs_2509_26476
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Regression Language Models for Code
Akhauri, Yash
Song, Xingyou
Wongpanich, Arissa
Lewandowski, Bryan
Abdelfattah, Mohamed S.
Computation and Language
Artificial Intelligence
Machine Learning
Performance
Software Engineering
We study code-to-metric regression: predicting numeric outcomes of code executions, a challenging task due to the open-ended nature of programming languages. While prior methods have resorted to heavy and domain-specific feature engineering, we show that a single unified Regression Language Model (RLM) using a frozen LLM encoder can simultaneously predict directly from text, (i) the memory footprint of code across multiple high-level languages such as Python and C++, (ii) the latency of Triton GPU kernels, and (iii) the accuracy and speed of trained neural networks represented in ONNX. In particular, a relatively small 300M parameter RLM based on T5Gemma, obtains $>$0.9 Spearman-rank on competitive programming submissions from APPS, and a single unified model achieves $>$0.5 average Spearman-rank across 17 separate languages from CodeNet. Furthermore, the RLM can obtain the highest average Kendall-Tau of 0.46 on five classic NAS design spaces previously dominated by graph neural networks, and simultaneously predict architecture latencies on numerous hardware platforms.
title Regression Language Models for Code
topic Computation and Language
Artificial Intelligence
Machine Learning
Performance
Software Engineering
url https://arxiv.org/abs/2509.26476