Forklift: An Extensible Neural Lifter
Not on the map. Why?
Abstract (arXiv)
The escalating demand to migrate legacy software across different Instruction Set Architectures (ISAs) has driven the development of assembly-to-assembly translators to map between their respective assembly languages. However, the development of these tools requires substantial engineering effort. State-of-the-art approaches use lifting, a technique where source assembly code is translated to an architecture-independent intermediate representation (IR) (for example, the LLVM IR) and use a pre-existing compiler to recompile the IR to the target ISA. However, the hand-written rules these lifters employ are sensitive to the particular compiler and optimization level used to generate the code and require significant engineering effort to support each new ISA. We propose Forklift, the first neural lifter that learns how to translate assembly to LLVM IR using a token-level encoder-decoder Transformer. We show how to incrementally add support to new ISAs by fine tuning the assembly encoder and freezing the IR decoder, improving the overall accuracy and efficiency. We collect millions of parallel LLVM IR, x86, ARM, and RISC-V programs across compilers and optimization levels to train Forklift and set up an input/output-based accuracy harness. We evaluate Forklift on two challenging benchmark suites and translate 2.5x more x86 programs than a state-of-the-art hand-written lifter and 4.4x more x86 programs than GPT-4 as well as enabling translation from new ISAs.
Key facts. Cited by 2 papers in the map (data through 7 Oct 2026). Builds on 16 of its 42 references that are in the map. First posted on arXiv 1 Apr 2024.
Builds on
16 of 42 references.
References · 16
- 2017Attention Is All You Needcited by 59,932
- 2020Language Models are Few-Shot Learnerscited by 32,199
- 2021Evaluating Large Language Models Trained on Codecited by 8,053
- 2019BART: Denoising Sequence-to-Sequence Pre-training for Natural Language Generation, Translation, and Comprehensioncited by 5,874
- 2015Neural Machine Translation of Rare Words with Subword Unitscited by 2,898
- 2018Subword Regularization: Improving Neural Network Translation Models with Multiple Subword Candidatescited by 632
- 2020Unsupervised Translation of Programming Languagescited by 262
- 2020Byte Pair Encoding is Suboptimal for Language Model Pretrainingcited by 189
- 2018Tree-to-tree Neural Networks for Program Translationcited by 93
- 2023Using Large Language Models to Generate JUnit Tests: An Empirical Studycited by 53
- 2022Code Translation with Compiler Representationscited by 42
- 2023SLaDe: A Portable Small Language Model Decompiler for Optimized Assemblycited by 28
4 more
- 2022Beyond the C: Retargetable Decompilation using Neural Machine Translationcited by 15
- 2019From Bilingual to Multilingual Neural Machine Translation by Incremental Trainingcited by 14
- 2020Training Multilingual Machine Translation by Alternately Freezing Language-Specific Encoders-Decoderscited by 8
- 2020Modeling Black-Box Components with Probabilistic Synthesiscited by 8
Built on by
Citing papers per year: 2026: 2.
2 papers.