Bibliographic record
Abstract
Multicore computers today are ubiquitous, from large distributed systems to desktops and mobile phones. Their complex memory architecture enables fast and efficient execution when used correctly. Many efforts have been invested in making the task of taking advantage of these architectures easier. Among them, the development of optimizing and parallelizing compilers as well as frameworks that provide skeletons for parallelism have abstracted the low-level details of the implementation and the optimization process. Techniques have been developed to optimize the efficiency of algorithms that fit particular skeletons, with one of the most commonly used of these being divide-and-conquer. However, the programmer still has to come up with the divide-and-conquer algorithm that fits such skeletons, often with an error-prone and tedious process. This research work aims at providing automated techniques to bridge the gap in the automated optimization process. Given an input program, our goal is to synthesize an equivalent divide-and-conquer implementation, which can then be optimized using the wealth of techniques available. The methods developed should be implementable in a practical tool leveraging the recent advances in program synthesis, but they should also act as a thinking tool for programmers when the solvers fail. This dissertation proposes methodologies to split the task of synthesizing a divide-and-conquer program into smaller but more tractable tasks. To solve these tasks, we propose both new deductive approaches to synthesize new functionality and an efficient use of existing synthesis methods. The combination of the methodologies and the synthesis techniques allows us to design a fully automated process, implemented in a tool Parsynt, which succeeds in synthesizing highly non-trivial divide-and-conquer algorithms from simple functions.
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 | 0.002 | 0.005 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.001 | 0.001 |
| Science and technology studies | 0.001 | 0.002 |
| Scholarly communication | 0.001 | 0.001 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.001 | 0.001 |
| 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".