A Comprehensive Study of Real-World Bugs in Machine Learning Model Optimization

Hao Guan, Ying Xiao, Jiaying Li, Yepang Liu, Guangdong Bai. 2023. International Conference on Software Engineering.

Read the paper · doi:10.1109/icse48619.2023.00024

What this paper does with SQLancer

Databases are named once, in the opening sentence, as another domain where optimisation over complex objects is error-prone -- alongside program compilers. The mention is reachable only through the citation marker and establishes an analogy rather than any relationship to SQLancer. An empirical study of real-world bugs in machine learning model optimization, characterising what goes wrong when a trained model is converted, quantised or otherwise optimised for deployment. Written by claude-opus-5 from the 1 place this paper refers to SQLancer. The quotations below are the paper's own words, stored verbatim when the text was extracted.

How it was classified

uses infrastructure — no

SQLancer is cited as related work; no reuse of its code, generator or workload is described.

extends technique — no

No SQLancer technique is extended; the citation is background.

compares with — no

No empirical comparison against SQLancer or one of its oracles is reported.

describes as state of the art — no

The text does not describe SQLancer as the state of the art.

SQLancer publications it cites (1)

Bibliography entries that resolved to a SQLancer publication, or to a paper by one of the project's authors. A sentence citing one of these numbers is a reference to SQLancer even when it never writes the name.

#	Entry	Matched as
23	M. Rigger and Z. Su, “Detecting optimization bugs in database engines via non-optimizing reference engine construction,” in Proceedings of the 28th ACM Joint Meeting on European Software Engineering Conference and Sym...	sqlancer publication · NOREC

Every place it refers to SQLancer (1)

1 sentence, each stored verbatim from the extracted text with where it was found and how. “Citation marker” means the sentence names no tool at all and was reached through a reference number that resolved to a SQLancer publication.

Id	Sentence	Found by	Where
M1	Similar to the optimization task in other software that handles complex objects, such as in program compilers [20]– [22] and databases [23], ML model optimization is an errorprone process.	citation marker background	I INTRODUCTION page 1

This page is rendered from _data/papers/paper_doi_10_1109_icse48619_2023_00024.json, extracted from supplied pdf. 12 pages, 54 references parsed.