Papers › An Empirical Study of Data Ability Boundary in LLMs' Math Reasoning

An Empirical Study of Data Ability Boundary in LLMs' Math Reasoning

23 Feb 2024arXiv:2403.00799archive 2025-07-28

Zui Chen, Yezeng Chen, Jiaqi Han, Zhijie Huang, Ji Qi, Yi Zhou

Large language models (LLMs) are displaying emergent abilities for math reasoning tasks,and there is a growing attention on enhancing the ability of open-source LLMs through supervised fine-tuning (SFT).In this paper, we aim to explore a general data strategy for supervised data to help optimize and expand math reasoning ability.Firstly, we determine the ability boundary of reasoning paths augmentation by identifying these paths' minimal optimal set.Secondly, we validate that different abilities of the model can be cumulatively enhanced by Mix of Minimal Optimal Sets of corresponding types of data, while our models MMOS achieve SOTA performance on series base models under much lower construction costs.Besides, we point out GSM-HARD is not really hard and today's LLMs no longer lack numerical robustness.Also, we provide an Auto Problem Generator for robustness testing and educational applications.Our code and data are publicly available at https://github.com/cyzhh/MMOS.

PaperPDFCodeCode Syntology ran

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):

Code

Syntology 11 code samples harvested from 1 repository linked to this paper: 9 ran (of which 0 constructed an object rather than computing a result; 5 with no instrument failure: 0 honoured, 2 violated, 3 with no contract checked; 4 where Syntology's instrument failed) · 2 unverified. Where Syntology's instrument failed: 4 where our contract draft was wrong; failures of Syntology's instrument, not of the code.

By repository: official repository: 11 samples from 1 repository, 9 ran (5 with no instrument failure; 4 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 on 2026-09-24 (UTC): the dates Syntology's graph wrote each sample's verification level, not when the code executed. 2 of the 9 samples that ran have no recorded date.

cyzhh/MMOS officialmentioned in papermentioned on GitHubSyntology: 11 samples, 9 ran (4 instrument-failed) · 2 unverified report

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

11 samples harvested: 9 ran (of which 0 constructed an object rather than computing a result; 5 with no instrument failure: 0 honoured, 2 violated, 3 with no contract checked; 4 where Syntology's instrument failed) · 2 unverified. “Where Syntology's instrument failed” counts failures of Syntology's instrument, not of the code. Syntology's record for this page last changed ; when this build read Syntology's graph is in the build record. That date is not when the samples ran.

2ran · violated contract
4ran · our draft was wrong
3ran
2unverified

Licence: 11 of the 11 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 cyzhh/MMOS. “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.

cyzhh/MMOS official repository 11 samples, 9 ran (4 instrument-failed) · 2 unverified
encode_with_messages_format in 21 papers train/finetune.py ran · our draft was wrong no licence file found · pointer only · bc1a96ec28a11a5f · report fixture: signature example:dict/0/?, tokenizer:callable/0/?, max_seq_length:int/0/int, mask_prompt:bool/0/bool · returned dict · level recorded, date not recorded
encode_with_prompt_completion_format in 25 papers train/finetune.py ran · our draft was wrong no licence file found · pointer only · 1906977026be9a86 · report fixture: signature example:dict/0/?, tokenizer:callable/0/?, max_seq_length:int/0/int, mask_prompt:bool/0/bool · returned dict · level recorded, date not recorded
is_digit in 33 papers src/grader.py ran · violated contract no licence file found · pointer only · 8ae8c583b21633b3 · report fixture: signature s:str/0/str · returned bool · level recorded 2026-09-24
majority_vote in 13 papers eval/evaluate.py ran · our draft was wrong fingerprinted no licence file found · pointer only · c6427c534ee50108 · report fixture: signature lst:list/1/int · returned int · level recorded 2026-09-24
math_equal src/grader.py ran no licence file found · pointer only · 65c738adb08e913c · report fixture: signature prediction:str/0/str, reference:str/0/str, include_percentage:bool/0/bool, is_close:bool/0/bool, timeout:bool/0/bool · returned bool · level recorded 2026-09-24
math_equal_process src/grader.py ran · violated contract fingerprinted no licence file found · pointer only · f63c83de2a26f67c · report fixture: signature param:list/1/str · returned bool · level recorded 2026-09-24
move_key_to_first src/rerank.py ran · our draft was wrong no licence file found · pointer only · 38ab3477602c199f · report fixture: signature dictionary:dict/0/?, key:str/0/str · returned OrderedDict · level recorded 2026-09-24
replace_var_names_in_code_blocks src/nodup.py ran no licence file found · pointer only · adff4d5b0a7ae14f · report fixture: signature text:str/0/str, args:MinimalArgs/0/? · returned str · level recorded 2026-09-24
show_example src/rerank.py ran no licence file found · pointer only · 2c87889db2b36bfb · report fixture: signature file_path:str/0/str, count:int/0/int · returned list · level recorded 2026-09-24
build_paths src/utils.py unverified no licence file found · pointer only · 65de7160ff3afee9 · report
replace_var_names src/nodup.py unverified no licence file found · pointer only · 74f23325e7b85336 · report

Tasks

Arithmetic ReasoningAutomated Theorem ProvingMath Word Problem Solving

Datasets

Introduced by this paper, per the archive.

MMOS

Results from the paper archive 2025-07-28

TaskDatasetModelMetricValueRank at snapshotLeaderboardReport
Arithmetic Reasoning GSM8K MMOS-DeepSeekMath-7B(0-shot,k=50) Accuracy 87.2 #35 of 164 Archive leaderboard report
Arithmetic Reasoning GSM8K MMOS-DeepSeekMath-7B(0-shot,k=50) Parameters (Billion) 7 #35 of 164 Archive leaderboard report
Arithmetic Reasoning GSM8K MMOS-DeepSeekMath-7B(0-shot) Accuracy 80.5 #72 of 164 Archive leaderboard report
Arithmetic Reasoning GSM8K MMOS-DeepSeekMath-7B(0-shot) Parameters (Billion) 7 #72 of 164 Archive leaderboard report
Arithmetic Reasoning GSM8K MMOS-CODE-34B(0-shot) Accuracy 80.4 #73 of 164 Archive leaderboard report
Arithmetic Reasoning GSM8K MMOS-CODE-34B(0-shot) Parameters (Billion) 34 #73 of 164 Archive leaderboard report
Arithmetic Reasoning GSM8K MMOS-CODE-7B(0-shot) Accuracy 73.9 #92 of 164 Archive leaderboard report
Arithmetic Reasoning GSM8K MMOS-CODE-7B(0-shot) Parameters (Billion) 7 #92 of 164 Archive leaderboard report
Automated Theorem Proving miniF2F-test MMOS-DeepSeekMath-7B ITP Lean #19 of 29 Archive leaderboard report
Automated Theorem Proving miniF2F-test MMOS-DeepSeekMath-7B Pass@1 28.3 #19 of 29 Archive leaderboard report
Automated Theorem Proving miniF2F-test MMOS-DeepSeekMath-7B cumulative 28.3 #19 of 29 Archive leaderboard report
Math Word Problem Solving ASDiv-A MMOS-DeepSeekMath-7B(0-shot) Execution Accuracy 87.6 #2 of 9 Archive leaderboard report
Math Word Problem Solving ASDiv-A MMOS-CODE-34B(0-shot) Execution Accuracy 85.1 #4 of 9 Archive leaderboard report
Math Word Problem Solving ASDiv-A MMOS-CODE-7B(0-shot) Execution Accuracy 78.6 #8 of 9 Archive leaderboard report
Math Word Problem Solving MATH MMOS-DeepSeekMath-7B(0-shot,k=50) Accuracy 63.7 #23 of 135 Archive leaderboard report
Math Word Problem Solving MATH MMOS-DeepSeekMath-7B(0-shot,k=50) Parameters (Billions) 7 #23 of 135 Archive leaderboard report
Math Word Problem Solving MATH MMOS-DeepSeekMath-7B(0-shot) Accuracy 55.0 #37 of 135 Archive leaderboard report
Math Word Problem Solving MATH MMOS-DeepSeekMath-7B(0-shot) Parameters (Billions) 7 #37 of 135 Archive leaderboard report
Math Word Problem Solving MATH MMOS-CODE-34B(0-shot) Accuracy 49.5 #50 of 135 Archive leaderboard report
Math Word Problem Solving MATH MMOS-CODE-34B(0-shot) Parameters (Billions) 34 #50 of 135 Archive leaderboard report
Math Word Problem Solving MATH MMOS-CODE-7B(0-shot) Accuracy 44.3 #69 of 135 Archive leaderboard report
Math Word Problem Solving MATH MMOS-CODE-7B(0-shot) Parameters (Billions) 7 #69 of 135 Archive leaderboard report
Math Word Problem Solving SVAMP MMOS-CODE-34B(0-shot) Execution Accuracy 80.6 #9 of 26 Archive leaderboard report
Math Word Problem Solving SVAMP MMOS-DeepSeekMath-7B(0-shot) Execution Accuracy 79.3 #10 of 26 Archive leaderboard report
Math Word Problem Solving SVAMP MMOS-CODE-7B(0-shot) Execution Accuracy 76.4 #11 of 26 Archive leaderboard report

Ranks are positions in the archive's leaderboards as they stood at the 2025-07-28 snapshot. Results published since then are not among these rows, so a rank here is not a current standing.

Methods

BASE

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