Project Foxtrot: Abstracting the Complexity of Bitstream Reverse Engineering Through ML
Bibliographic record
Abstract
Field-programmable gate arrays (FPGAs) face substantial security threats due to their inherent reconfigurability, heavy reliance on third-party intellectual property (3PIP) blocks, and increasing deployment in safety-critical domains. To compound these challenges, the proprietary nature of FPGA bitstreams hinders the ability to verify designs against potential threats such as hardware Trojans. Given these security challenges, prior works have proposed reverse engineering (RE) solutions to recover netlist information from bitstreams for security analysis. However, these solutions are FPGA-specific and require an exhaustive mapping between the netlist functions and bitstream bits. This paper introduces Project Foxtrot, a novel, vendor-agnostic framework for FPGA bitstream RE. Our framework leverages machine learning (ML) to simplify and accelerate the bitstream mapping process. It addresses the limitations of previous works by providing a universal approach that can be adapted to various FPGA vendors. The work presented focuses on the ML techniques integrated within our framework. We demonstrate how these techniques are used to abstract complex, time-consuming tasks, thereby making the process more efficient and scalable. We present our framework and demonstrate the effectiveness of our approach using a lookup table (LUT) recovery use case. Our experiments, conducted on the Intel-Altera Flex 10K, the AMD-Xilinx Spartan 7, and the AMD-Xilinx Artix 7 FPGAs, show perfect accuracy in LUT truth table recovery, surpassing recent state-of-the-art approaches.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.001 | 0.000 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.000 |
| Bibliometrics | 0.000 | 0.002 |
| Science and technology studies | 0.001 | 0.001 |
| Scholarly communication | 0.000 | 0.001 |
| Open science | 0.002 | 0.001 |
| Research integrity | 0.000 | 0.001 |
| Insufficient payload (model declined to judge) | 0.000 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one teacher head, not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".