Enabling Language-Specific Transformations in Language-Agnostic Program Reduction
Bibliographic record
Abstract
When a program P triggers a bug in a language implementation, program reduction can reduce P by removing program elements that are irrelevant to the bug, to facilitate debugging. Program reduction has been widely used in communities of various language implementations. Generally, program reduction techniques can be classified into language-agnostic program reducers (ARs) category and language-specific program reducers (SRs) category. ARs work generally well in a wide range of languages but usually produce less optimal results than SRs due to a lack of domain knowledge of specific languages. However, SRs require extensive engineering effort to leverage the domain knowledge, and can only function in their target language but not in other languages. \nTo combine the benefits of both ARs and SRs and minimize the gap between the two, a novel, general transformation framework, Metis,1 is introduced. Specifically, Metis allows users to specify language-specific program transformations to further minimize the results by SRs and the users only need to know the syntax of the target language and a concise domain-specific language named MTL (Metis Transformation Language) provided by Metis; Metis automatically processes the transformation rules inscribed in MTL by performing pattern matching and subsequent rewriting operations on the parse tree of the program under reduction. Metis provides a general, unified framework for specifying program transformations for different languages. \nWe comprehensively evaluated Metis on two benchmark sets of C and SMT-LIB pro- grams and the results demonstrate that Metis yields much smaller programs than the state-of-the-art language-agnostic program reducer by 35.8% on average. We also compared Metis with two SRs: ddSMT and C-Reduce. Metis produces results of comparable size to ddSMT, but with a noticeable 28.9% shorter reduction time; while falling short of matching the reduced program size by C-Reduce, Metis saves 82.4% of queries and achieves a speed improvement of 30.6% less runtime.
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 machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.003 | 0.008 |
| Meta-epidemiology (narrow) | 0.002 | 0.001 |
| Meta-epidemiology (broad) | 0.001 | 0.002 |
| Bibliometrics | 0.002 | 0.001 |
| Science and technology studies | 0.001 | 0.002 |
| Scholarly communication | 0.002 | 0.004 |
| Open science | 0.003 | 0.003 |
| Research integrity | 0.001 | 0.003 |
| Insufficient payload (model declined to judge) | 0.003 | 0.001 |
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 source (direct Gemma or distilled Codex), 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".