Papers › Learning Performance-Improving Code Edits
Learning Performance-Improving Code Edits
Alexander Shypula, Aman Madaan, Yimeng Zeng, Uri Alon, Jacob Gardner, Milad Hashemi, Graham Neubig, Parthasarathy Ranganathan, Osbert Bastani, Amir Yazdanbakhsh
With the decline of Moore's law, optimizing program performance has become a major focus of software research. However, high-level optimizations such as API and algorithm changes remain elusive due to the difficulty of understanding the semantics of code. Simultaneously, pretrained large language models (LLMs) have demonstrated strong capabilities at solving a wide range of programming tasks. To that end, we introduce a framework for adapting LLMs to high-level program optimization. First, we curate a dataset of performance-improving edits made by human programmers of over 77,000 competitive C++ programming submission pairs, accompanied by extensive unit tests. A major challenge is the significant variability of measuring performance on commodity hardware, which can lead to spurious "improvements." To isolate and reliably evaluate the impact of program optimizations, we design an environment based on the gem5 full system simulator, the de facto simulator used in academia and industry. Next, we propose a broad range of adaptation strategies for code optimization; for prompting, these include retrieval-based few-shot prompting and chain-of-thought, and for finetuning, these include performance-conditioned generation and synthetic data augmentation based on self-play. A combination of these techniques achieves a mean speedup of 6.86 with eight generations, higher than average optimizations from individual programmers (3.66). Using our model's fastest generations, we set a new upper limit on the fastest speedup possible for our dataset at 9.64 compared to using the fastest human submissions available (9.56).
In Syntology View this paper on Syntology, its page in Syntology's graph. That page adds which of its functions are also present in other papers' code Syntology holds, with dates, and its dated run record as JSON and a README badge. It also lists the repositories linked to the paper, the abstract and the calls for agents.
Open this paper in Syntology's Atlas, the map of the papers in Syntology's graph and their citations.
For agents, on Syntology's MCP service (how to connect):
get_harvested_code_for_paper(arxiv_id="2302.07867")get_code(code_sha256="…")with a sample'scode_sha256from the list belowget_citation_path(paper_1="2302.07867", paper_2="…")with another paper's arXiv id or titleget_concepts_for_paper(arxiv_id="2302.07867")
Code
Syntology 19 code samples harvested from 1 repository linked to this paper: 8 ran (of which 0 constructed an object rather than computing a result; 3 with no instrument failure: 2 honoured, 0 violated, 1 with no contract checked; 5 where Syntology's instrument failed) · 11 unverified. Where Syntology's instrument failed: 5 where our contract draft was wrong; failures of Syntology's instrument, not of the code.
By repository: official repository: 19 samples from 1 repository, 8 ran (3 with no instrument failure; 5 where Syntology's instrument failed). The run record, sample by sample. “Ran” means executed on a synthesized input, not that the code is correct or reproduces the paper.
Levels written between 2026-09-22 and 2026-09-25 (UTC): the dates Syntology's graph wrote each sample's verification level, not when the code executed.
Repository list and official/mentioned flags are the archive's, frozen 2025-07-28. Reachability, where shown, is from one Syntology probe window (2026-09-16 to 2026-09-18); repositories not probed show nothing. GitHub stars are not tracked.
Code Syntology ran Syntology
19 samples harvested: 8 ran (of which 0 constructed an object rather than computing a result; 3 with no instrument failure: 2 honoured, 0 violated, 1 with no contract checked; 5 where Syntology's instrument failed) · 11 unverified. “Where Syntology's instrument failed” counts failures of Syntology's instrument, not of the code. Syntology's record for this page has not changed since , the first build that kept a record date for it; when this build read Syntology's graph is in the build record. That date is not when the samples ran.
Licence: 19 of the 19 samples are pointer only, meaning Syntology does not serve that copy's text. This page shows no code text for any sample; each one links to its file in the repository.
Harvested from madaan/pie-perf. “Ran” means the sample executed on a synthesized input. It does not mean the output is correct, and nothing here reproduces the paper's results. A line beginning “fixture:” under a sample that ran gives the arguments Syntology's fixture called the sample with, each as name:type/rank/element kind (? where none was recorded), and the type of what the fixture's run returned, exactly as the harness recorded them; the arguments and the returned type were recorded by two separate runs of the same fixture. No output value is shown, and a returned type says nothing about whether the output was right. “Level recorded” gives the UTC date when Syntology's graph wrote the sample's verification level, not when the code executed; where the level was written before Syntology's graph kept that date, the row says “level recorded, date not recorded”. “Honoured” and “violated” refer to a contract Syntology drafted from the code itself. “Our draft was wrong” and “the contract check could not run” are failures of Syntology's instrument, not of the code: the sample ran on Syntology's synthesized input, and then the drafted contract did not describe it, or the check itself raised an error (RAISES), lacked a dependency (DEP_MISSING) or ran out of time (TIMEOUT); hover the label for which. A metamorphic tier, where shown, is evidence of the output's shape, not that the sample implements the method.
Each sample ends with its code_sha256. A sample's code_sha256 is Syntology's identity for that exact code. It is printed in full (16 hexadecimal characters for every sample in this build), and get_code(code_sha256="…") takes it exactly as printed. An agent calls it on Syntology's MCP service (how to connect); click an identity to copy that call.
Repository labels, per sample. official repository: The archive marks this repository official for the paper. named in the paper: The archive records that the paper mentions this repository; it is not marked official. community (archive-listed): In the archive's code links for this paper, not marked official and not recorded as mentioned in the paper. found in paper text by Syntology: Syntology found this repository in the paper's own text; whether it is the authors' implementation is not asserted. community: Not in the archive's code links for this paper; a community repository Syntology harvested. Samples are grouped by repository: a repository marked official first, then by how directly the repository is tied to the paper, then in the archive's order; never by how many ran. Licence labels name the repository's licence as recorded at harvest. “Pointer only” means Syntology does not serve that copy's text, for one of six reasons: no licence file was found; the licence was not identified; the licence is recorded as permissive but that copy's record is not marked cleared; the licence is outside the permissive list Syntology serves text under (MIT, Apache-2.0, BSD and similar); this paper's copy was not located, so its licence is not recorded; or this paper's copy is cleared, but the stored copy of the same code that Syntology serves is not. Some licences outside that list permit redistribution, such as WTFPL, and GPL-3.0 under its conditions; they are simply not on the list. Hover a licence label for the reason. File links open the file on GitHub at the default branch, which may have changed since the harvest.
A function or class name links to its page when the name occurs in the harvested code of at least 5 papers; the name is a string, not an identity. A census tag appears when Syntology's census ran implementations with that name on one shared input per bucket: it gives how many were compared, from how many papers, and how many distinct outputs they produced, and names this sample's bucket when this sample was one of them.
540fd96572641577 · report
fixture: signature run_metrics:DataFrame/0/? · returned DataFrame · level recorded 2026-09-22 aef90ef51a665bd5 · report
fixture: signature results:dict/0/?, gen_col:str/0/str · returned dict · level recorded 2026-09-24 52f66cb1672a41ff · report
fixture: signature row:Series/1/str, question_sep:str/0/str, answer_sep:str/0/str · returned str · level recorded 2026-09-25 4e1a10fed322555f · report
fixture: returned bool · level recorded 2026-09-22 c101803bbb9b31c7 · report
fixture: signature df:DataFrame/0/?, col:str/0/str · returned str · level recorded 2026-09-24 7eaeae08022b7013 · report
fixture: signature args:list/1/str, input_file_path:str/0/str, timeout_seconds:int/0/int · returned str · level recorded 2026-09-24 773a60b768feb034 · report
fixture: signature code_path:str/0/str, unit_test_data_basepath:str/0/str, num_runs_per_test_case:int/0/int, ignore_first_k:int/0/int, max_seconds_per_run:int/0/int, ground_truths:list/1/str, num_test_cases:int/0/int, cpu_number:int/0/int, return_per_trial_times:bool/0/bool, python_bin:str/0/str, return_dict:bool/0/bool, remove_code_after_run:bool/0/bool, cflags:str/0/str, return_if_acc_below:float/0/float · returned tuple · level recorded 2026-09-24 b14554bb5b015337 · report
fixture: signature code_path:str/0/str, unit_test_data_basepath:str/0/str, num_runs_per_test_case:int/0/int, ignore_first_k:int/0/int, max_seconds_per_run:int/0/int, ground_truths:list/1/str, num_test_cases:int/0/int, cpu_number:int/0/int, return_per_trial_times:bool/0/bool, python_bin:str/0/str, return_dict:bool/0/bool, cflags:NoneType/0/?, return_if_acc_below:NoneType/0/? · returned tuple · level recorded 2026-09-24 37991ff1b0138b16 · report
ed1e7b9c5b8615ca · report
c72a3da727492d69 · report
8b19025c6c05d32c · report
census: 4 implementations from 6 papers, 4 distinct outputs across 4 buckets, one shared input per bucket src/codenet_eval/run_eval.py unverified no licence file found · pointer only ·
1f09a6f0ad0d135a · report
3a991d80b8cbece2 · report
77ca852b3aaaa81a · report
7d44d630c645c37e · report
973eee46109fd821 · report
6db030a5cc638ec1 · report
a177ad6c3baf7d7a · report
Method as code Syntology
What Syntology's graph records as this paper's claimed method, lifted into code: each assembly names its bricks (functions and classes), each with its own role and its own verification level. An assembly is not a check that the code implements the method correctly or reproduces the paper's results. Levels are each brick's own verification level in Syntology's graph (level 0: no recorded run), listed brick by brick and never combined into one level; “ran” means the brick executed on a synthesized input, not that its output is correct, and the instrument figure counts failures of Syntology's instrument, not of the code. Where a model read the code against the paper's claim, each assembly's label verdict is one model's judgement, kept in this page's JSON twin; it is not calibrated, so no card prints it.
1 assembly on this paper: 1 harvested, 0 generated.
harvested Assembly for framework for adapting LLMs to high-level program optimization (graph method: framework for adapting LLMs to high-level program optimization). 19 bricks; individually on a fixture: 8 ran (of which 0 constructed an object rather than computing a result; 3 with no instrument failure: 2 honoured, 0 violated, 1 with no contract checked; 5 where Syntology's instrument failed) · 11 unverified. The assembly has not been run end-to-end.
- evaluate_generated_outputs entry · level 0 · unverified ·
8b19025c6c05d32c - EvaluationConfig helper · level 0 · unverified ·
37991ff1b0138b16 - get_best_generation_per_submission metric · level 1 · ran · our draft was wrong ·
aef90ef51a665bd5 - print_summary helper · level 0 · unverified ·
77ca852b3aaaa81a - read_inputs_and_prepare helper · level 0 · unverified ·
7d44d630c645c37e - run_programs metric · level 0 · unverified ·
6db030a5cc638ec1 - write_programs_read_ground_truth helper · level 0 · unverified ·
a177ad6c3baf7d7a - get_anomalies metric · level 1 · ran · our draft was wrong ·
540fd96572641577 - mean_std helper · level 1 · ran · our draft was wrong ·
c101803bbb9b31c7 - _prepare_for_rerun helper · level 0 · unverified ·
ed1e7b9c5b8615ca - get_input_from_prompt data_transform · level 1 · ran · our draft was wrong ·
52f66cb1672a41ff - run_code_on_inputs helper · level 0 · unverified ·
973eee46109fd821 - run_cpp_code_on_inputs helper · level 1 · ran · honoured contract ·
773a60b768feb034 - run_python_code_on_inputs helper · level 1 · ran · honoured contract ·
b14554bb5b015337 - compile_cpp_code helper · level 0 · unverified ·
c72a3da727492d69 - get_accuracy helper · level 0 · unverified ·
1f09a6f0ad0d135a - is_linux helper · level 1 · ran ·
4e1a10fed322555f - run_cmd_for_time_eval helper · level 1 · ran · our draft was wrong ·
7eaeae08022b7013 - limit_virtual_memory helper · level 0 · unverified ·
3a991d80b8cbece2
Each brick ends with its code_sha256, Syntology's identity for that exact code; click one to copy the get_code call that returns it. For agents, on Syntology's MCP service: assembly(arxiv_id="2302.07867") returns these assemblies with each brick's role and wiring (how to connect).
Tasks
Datasets
Introduced by this paper, per the archive.
Results from the paper archive 2025-07-28
No leaderboard rows for this paper in the archive.
Methods
Report a problem or propose a change · a person checks every report against the paper or source before anything changes; decisions are listed on /corrections