← 研踪 PaperTrailarXiv:2610.36288 · v1
BUSTA: Teaching Small Language Models to Reason About Binary Execution
首次提交 2026-10-27 05:11
版本更新 2026-10-27 05:11
雷达观测 2026-10-29 00:35
摘要
Large Language Models (LLM) are increasingly applied to Reverse Engineering (RE) workflows. RE, however, differs from common LLM coding tasks as it requires models to reason about low-level program behavior. Furthermore, the deployment of tooling in resource-constrained sandboxed RE environments limits the ability to use frontier-scale solutions. We examine whether a targeted LLM training approach can transfer distinct RE skills to open models via inverse basic-block execution: given an x86-64 basic block and a desired output register state, the model predicts the input register values that produce it. Fine-tuning \texttt{Qwen3-1.7B} and \texttt{Nemotron-3-Nano} with Group Relative Policy Optimization (GRPO) raises knowable-register prediction accuracy from 73.2\% to 85.3\% and from 63.3\% to 95.2\%, respectively. Frontier models still score highest (up to 99.2\%), but our fine-tuned \texttt{Nemotron-3-Nano} comes within four points at a fraction of the parameter count. We release both fine-tuned models and the training dataset.
观测到的版本
相邻记录可能跳过中间版本；首次发现 v2 及以上也会标为版本更新。
v1 · 2026-10-27 05:11 · 最新观测版本
BUSTA: Teaching Small Language Models to Reason About Binary Execution
Large Language Models (LLM) are increasingly applied to Reverse Engineering (RE) workflows. RE, however, differs from common LLM coding tasks as it requires models to reason about low-level program behavior. Furthermore, the deployment of tooling in resource-constrained sandboxed RE environments limits the ability to use frontier-scale solutions. We examine whether a targeted LLM training approach can transfer distinct RE skills to open models via inverse basic-block execution: given an x86-64 basic block and a desired output register state, the model predicts the input register values that produce it. Fine-tuning \texttt{Qwen3-1.7B} and \texttt{Nemotron-3-Nano} with Group Relative Policy Optimization (GRPO) raises knowable-register prediction accuracy from 73.2\% to 85.3\% and from 63.3\% to 95.2\%, respectively. Frontier models still score highest (up to 99.2\%), but our fine-tuned \texttt{Nemotron-3-Nano} comes within four points at a fraction of the parameter count. We release both fine-tuned models and the training dataset.
查看此版本解读 →查看此版本原文 ↗