ASMA-Tune: Unlocking LLMs' Assembly Code Comprehension via Structural-Semantic Instruction Tuning

Assembly code analysis and comprehension play critical roles in applications like reverse engineering, yet they face substantial challenges due to low information density and a lack of explicit syntactic structures. While traditional masked language modeling (MLM) approaches do not explicitly focus on natural language interaction, emerging decoder-focused large language models (LLMs) demonstrate partial success in binary analysis yet remain underexplored for holistic comprehension. We present Assembly Augmented Tuning, an end-to-end structural-semantic instruction tuning framework that synergizes encoder architecture with decoder-based LLMs through a projector module, where the assembly encoder extracts hardware-level structural features, the projector bridges representations with the semantic space, and the instruction-tuned LLM preserves natural language capabilities. Experimental results demonstrate three key advantages: (1) State-of-the-art performance in assembly comprehension with +39.7% Recall@1 and +17.8% MRR improvements over GPT-4-Turbo, (2) Consistent enhancements across base models (24.6-107.4% Recall@1 and 15.2-106.3% MRR on Qwen2.5-Coder, Deepseek-Coder and CodeLlama variants), and (3) Superior instruction-following capabilities (41.5%-118% improvements) with controlled code generation degradation (-8.9% to -35% across architectures).

@article{wang2025_2503.11617,
  title={ ASMA-Tune: Unlocking LLMs' Assembly Code Comprehension via Structural-Semantic Instruction Tuning },
  author={ Xinyi Wang and Jiashui Wang and Jinbo Su and Ke Wang and Peng Chen and Yanming Liu and Long Liu and Xiang Li and Yangdong Wang and Qiyuan Chen and Rongze Chen and Chunfu Jia },
  journal={arXiv preprint arXiv:2503.11617},
  year={ 2025 }
}