PathDiff: Systematic Differential Testing Using Symbolic Analysis
Why this work is in the frame
A frame that forgets how it found something cannot be audited. These are the routes that admitted this work.
Bibliographic record
Abstract
Discrepancies among software programs that implement the same specification frequently lead to bugs and vulnerabilities. Differential fuzzers, such as NEZHA, increase the efficiency of finding discrepancies by guiding the fuzzing process with domain-independent behavioural asymmetries. However, the random nature of fuzzing inevitably causes NEZHA to miss discrepancies, even if the pair of paths where the discrepancies lie has already been discovered. In this thesis, we propose PathDiff, a novel differential testing tool that leverages path asymmetry to systematically find all discrepancies for given path-pairs. PathDiff keeps the execution path in one program and iteratively negates each branch in the other program for discrepancies, exhaustively enumerating all discrepancies for given inputs. We have implemented PathDiff and evaluated it against NEZHA on 16 applications, including 12 applications used in the NEZHA paper as well as 4 newly selected ones. The results show that PathDiff finds 4.1× more discrepancies than NEZHA.
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.
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 | 0.003 |
| Meta-epidemiology (narrow) | 0.001 | 0.001 |
| Meta-epidemiology (broad) | 0.002 | 0.001 |
| Bibliometrics | 0.002 | 0.008 |
| Science and technology studies | 0.002 | 0.000 |
| Scholarly communication | 0.001 | 0.000 |
| Open science | 0.003 | 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 it