# ICALL-BENCH: Benchmarking LLM Interprocedural Reasoning in Binary Programs

> ICALL-BENCH measures whether language models and agents can follow interprocedural meaning in compiled binaries, using indirect-call target recovery as the probe. The abstract says the benchmark covers 50 real-world C and C++ programs built with different compilers.

- Authors: Shenglong Yao; Zirui Liu; Xiangchi Yuan; Wenke Lee
- Venue: arXiv
- Status: Preprint
- Publication / posting date: October 27, 2026
- Date source: arXiv first-submission date; the cited version is v1
- First arXiv submission: October 27, 2026
- arXiv: 2610.37871
- Reviewed arXiv version: v1
- Sources reviewed: 2026-11-04
- Added to the catalog: 2026-11-09
- Original paper: <https://arxiv.org/abs/2610.37871v1>
- PDF: <https://arxiv.org/pdf/2610.37871v1>
- BibTeX: <https://awesomerea.org/papers/citations/paper-icall-bench.bib>
- Catalog page: <https://awesomerea.org/papers/paper-icall-bench/>
- Record JSON: <https://awesomerea.org/papers/paper-icall-bench/record.json>
- Chinese version: [简体中文](https://awesomerea.org/zh/papers/paper-icall-bench/index.md)

## Scope and limitations

arXiv:2610.37871v1 reports a benchmark of 50 real-world programs.

## Why it fits this collection

Recovering indirect-call targets is a binary analysis task, and the paper contributes a benchmark for it.

## Publication and review notes

No code commit was pinned.

## Primary sources

1. <https://arxiv.org/abs/2610.37871v1>
2. <https://arxiv.org/pdf/2610.37871v1>
3. <https://export.arxiv.org/api/query?id_list=2610.37871>

## About this entry

Reviewed against the paper and its primary sources. Cite the original authors using the linked paper or BibTeX.
Original editorial content: MIT. Linked projects, papers and assets retain their own rights and licenses.