Papers › DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Models

DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Models

5 Feb 2024arXiv:2402.03300archive 2025-07-28

Zhihong Shao, Peiyi Wang, Qihao Zhu, Runxin Xu, Junxiao Song, Xiao Bi, Haowei Zhang, Mingchuan Zhang, Y. K. Li, Y. Wu, Daya Guo

Mathematical reasoning poses a significant challenge for language models due to its complex and structured nature. In this paper, we introduce DeepSeekMath 7B, which continues pre-training DeepSeek-Coder-Base-v1.5 7B with 120B math-related tokens sourced from Common Crawl, together with natural language and code data. DeepSeekMath 7B has achieved an impressive score of 51.7% on the competition-level MATH benchmark without relying on external toolkits and voting techniques, approaching the performance level of Gemini-Ultra and GPT-4. Self-consistency over 64 samples from DeepSeekMath 7B achieves 60.9% on MATH. The mathematical reasoning capability of DeepSeekMath is attributed to two key factors: First, we harness the significant potential of publicly available web data through a meticulously engineered data selection pipeline. Second, we introduce Group Relative Policy Optimization (GRPO), a variant of Proximal Policy Optimization (PPO), that enhances mathematical reasoning abilities while concurrently optimizing the memory usage of PPO.

PaperPDFCodeCode Syntology ranMethod as code

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 24 code samples harvested from 3 repositories linked to this paper: 14 ran (of which 0 constructed an object rather than computing a result; 3 with no instrument failure: 0 honoured, 0 violated, 3 with no contract checked; 11 where Syntology's instrument failed) · 10 unverified. Where Syntology's instrument failed: 9 where our contract draft was wrong; 2 where the contract check could not run; failures of Syntology's instrument, not of the code.

By repository: official repository: 9 samples from 1 repository, 8 ran (1 with no instrument failure; 7 where Syntology's instrument failed); community (archive-listed): 15 samples from 2 repositories, 6 ran (2 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.

No sample is linked to this paper from 2 of its repositories that Syntology harvested for other papers:

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. 3 of the 14 samples that ran have no recorded date.

deepseek-ai/deepseek-math officialmentioned in papermentioned on GitHubpytorchMITSyntology: 9 samples, 8 ran (7 instrument-failed) · 1 unverified report
openpsi-project/realhf mentioned on GitHubpytorchSyntology: 1 sample, 1 ran (1 instrument-failed) · 0 unverified report
shibing624/medicalgpt mentioned on GitHubpytorchApache-2.0Syntology: no sample linked to this paper (harvested for another paper). Syntology's graph links these samples to ORPO: Monolithic Preference Optimization without Reference Model (arXiv:2403.07691; that paper's own run record: 11 ran (of which 1 constructed an object rather than computing a result; 9 with no instrument failure: 1 honoured, 0 violated, 8 with no contract checked; 2 where Syntology's instrument failed) · 3 unverified); nothing checks that they implement this paper's method. report
synthlabsai/big-math mentioned on GitHubMITSyntology: no sample linked to this paper (harvested for another paper). Syntology's graph links these samples to Big-Math: A Large-Scale, High-Quality Math Dataset for Reinforcement Learning in... (arXiv:2502.17387; that paper's own run record: 14 ran (of which 0 constructed an object rather than computing a result; 14 with no instrument failure: 0 honoured, 0 violated, 14 with no contract checked; 0 where Syntology's instrument failed) · 1 unverified); nothing checks that they implement this paper's method. report
yynil/rwkvinside mentioned on GitHubpytorchApache-2.0Syntology: 14 samples, 5 ran (3 instrument-failed) · 9 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.

A paper named beside a repository with no sample linked to this paper is shown with that paper's own run record, not this paper's: “ran” means executed on a synthesized input, not that the code is correct, and “where Syntology's instrument failed” counts failures of Syntology's instrument, not of the code. Papers are listed in arXiv-id order, at most three per repository.

Code Syntology ran Syntology

24 samples harvested: 14 ran (of which 0 constructed an object rather than computing a result; 3 with no instrument failure: 0 honoured, 0 violated, 3 with no contract checked; 11 where Syntology's instrument failed) · 10 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.

9ran · our draft was wrong
2ran; the contract check could not run
3ran
10unverified

Licence: 3 of the 24 samples are pointer only, meaning Syntology does not serve that copy's text. For 3 of them, this paper's copy is under a permissive licence and cleared, but the stored copy of the same code that Syntology serves is not; the licence label's tooltip gives Syntology's recorded reason. This page shows no code text for any sample; each one links to its file in the repository.

Harvested from 3 repositories linked to this paper, official or community; the list below is grouped by repository, each group headed by its name and what it is to the paper. “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.

deepseek-ai/deepseek-math official repository 9 samples, 8 ran (7 instrument-failed) · 1 unverified
aggregate_metrics in 13 papers evaluation/summarize_results.py ran · our draft was wrong MIT (permissive) · 8b9e2baf73bb9a6f · report fixture: signature paths:list/1/str · returned dict · level recorded 2026-09-25
aggregate_predictions in 5 papers evaluation/summarize_results.py ran · our draft was wrong fingerprinted MIT (permissive) · 33425041e7f88b7f · report fixture: signature paths:list/1/str · returned list · level recorded 2026-09-25
extract_boxed_answers in 9 papers evaluation/data_processing/answer_extraction.py ran · our draft was wrong fingerprinted MIT (permissive) here; the stored copy is pointer only · 6390dea763ce122e · report fixture: signature text:str/0/str · returned list · level recorded 2026-09-24
extract_program_output in 6 papers evaluation/data_processing/answer_extraction.py ran · our draft was wrong fingerprinted MIT (permissive) here; the stored copy is pointer only · 0772e5020f6b0849 · report fixture: signature pred_str:str/0/str · returned str · level recorded 2026-09-24
markup_question evaluation/run_subset_parallel.py ran MIT (permissive) · de0231790ab22287 · report fixture: signature args:NoneType/0/?, item:dict/0/?, language:str/0/str, src:str/0/str, task:str/0/str · returned dict · level recorded 2026-09-25
read_data in 132 papers evaluation/utils.py ran · our draft was wrong MIT (permissive) · 3deb39de933b985a · report fixture: signature path:str/0/str · returned list · level recorded 2026-09-25
shuffle in 35 papers evaluation/utils.py ran · our draft was wrong fingerprinted MIT (permissive) · 1d4da55afb0377ec · report fixture: signature data:list/1/str, seed:int/0/int · returned list · level recorded 2026-09-24
strip_string in 18 papers evaluation/data_processing/answer_extraction.py ran · our draft was wrong fingerprinted MIT (permissive) here; the stored copy is pointer only · 5abed389274c4c03 · report fixture: signature string:str/0/str · returned str · level recorded 2026-09-24
check_proof evaluation/unsafe_score_minif2f_isabelle.py unverified MIT (permissive) · c59c51644b25a189 · report
openpsi-project/realhf community (archive-listed) 1 sample, 1 ran (1 instrument-failed) · 0 unverified
actor_loss_fn realhf/impl/model/utils/ppo_functional.py ran; the contract check could not run (RAISES) · metamorphic tier: deterministic Apache-2.0 (permissive) · e2e940c65c7460d3 · report fixture: signature logprobs:Tensor/2/float, old_logprobs:Tensor/2/float, advantages:Tensor/2/float, eps_clip:float/0/float, loss_mask:Tensor/2/bool · returned tuple · level recorded 2026-09-22
yynil/rwkvinside community (archive-listed) 14 samples, 5 ran (3 instrument-failed) · 9 unverified
repeat_kv in 244 papers
census: 15 implementations from 60 papers, 2 distinct outputs on one shared input this sample: bucket 1 of 1, arg 1: rank 4, kind float, dtype float32
rwkv_inside/modeling_arwkv.py ran; the contract check could not run (RAISES) · metamorphic tier: invariant fingerprinted Apache-2.0 (permissive) · 30d7eec482ebf6b1 · report fixture: signature hidden_states:Tensor/4/float, n_rep:int/0/int · returned Tensor · level recorded, date not recorded
apply_rotary_pos_emb in 393 papers
census: 14 implementations from 15 papers, 8 distinct outputs across 7 buckets, one shared input per bucket
rwkv_inside/modeling_arwkv.py ran · our draft was wrong Apache-2.0 (permissive) · bac65c3dafaec040 · report fixture: signature q:Tensor/4/float, k:Tensor/4/float, cos:Tensor/3/float, sin:Tensor/3/float, position_ids:NoneType/0/?, unsqueeze_dim:int/0/int · returned tuple · level recorded, date not recorded
compare_latex_numbers rl/utilities.py ran fingerprinted Apache-2.0 (permissive) · 620dc092b02e33f7 · report fixture: signature expr1:str/0/str, expr2:str/0/str · returned bool · level recorded 2026-09-24
parse_coordinate rl/utilities.py ran fingerprinted Apache-2.0 (permissive) · eb7f513a0e05ec8b · report fixture: signature coord_str:str/0/str · returned tuple · level recorded 2026-09-24
rotate_half in 424 papers
census: 24 implementations from 96 papers, 4 distinct outputs across 2 buckets, one shared input per bucket this sample: bucket 1 of 2, arg 1: rank 2, kind float, dtype float32
rwkv_inside/modeling_arwkv.py ran · our draft was wrong fingerprinted Apache-2.0 (permissive) · b99eea6376d1e212 · report fixture: signature x:Tensor/2/float · returned Tensor · level recorded, date not recorded
calculate_language_consistency_score rl/deepspeed_rl_train.py unverified Apache-2.0 (permissive) · 8ddbdfd0041abedc · report
calculate_param_sizes train_scripts/split_model_params.py unverified Apache-2.0 (permissive) · 93678d5fdf25571d · report
calculate_tensor_size train_scripts/split_model_params.py unverified Apache-2.0 (permissive) · bc474830fb3ccd5d · report
distribute_params train_scripts/split_model_params.py unverified Apache-2.0 (permissive) · d0a5e652d9881e59 · report
grpo_loss_with_old_logps rl/grpo_trainer.py unverified Apache-2.0 (permissive) · f9c058ca4549e9f6 · report
parse_latex_to_python rl/utilities.py unverified Apache-2.0 (permissive) · f7cc3a0c10540f2e · report
reward_function in 7 papers rl/train_rl_model.py unverified Apache-2.0 (permissive) · f872aa95062d3ac2 · report
selective_log_softmax_old rl/grpo_trainer.py unverified Apache-2.0 (permissive) · 7c05d945015e9e96 · report
time_function in 7 papers rl/profiler.py unverified Apache-2.0 (permissive) · b016e9b55d3c8dad · 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 Group Relative Policy Optimization (GRPO) (graph method: Group Relative Policy Optimization (GRPO)). 1 brick; individually on a fixture: 1 ran (of which 0 constructed an object rather than computing a result; 0 with no instrument failure: 0 honoured, 0 violated, 0 with no contract checked; 1 where Syntology's instrument failed) · 0 unverified. The assembly has not been run end-to-end.

  1. actor_loss_fn entry · level 1 · ran; the contract check could not run (RAISES) · e2e940c65c7460d3

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="2402.03300") returns these assemblies with each brick's role and wiring (how to connect).

Tasks

Arithmetic ReasoningMathMath Word Problem SolvingMathematical Reasoning

Results from the paper archive 2025-07-28

TaskDatasetModelMetricValueRank at snapshotLeaderboardReport
Arithmetic Reasoning GSM8K DeepSeekMATH-RL-7B Accuracy 88.2 #29 of 164 Archive leaderboard report
Arithmetic Reasoning GSM8K DeepSeekMATH-RL-7B Parameters (Billion) 7 #29 of 164 Archive leaderboard report
Math Word Problem Solving MATH DeepSeekMATH-RL-7B (w/ code, greedy decoding) Accuracy 58.8 #28 of 135 Archive leaderboard report
Math Word Problem Solving MATH DeepSeekMATH-RL-7B (w/ code, greedy decoding) Parameters (Billions) 7 #28 of 135 Archive leaderboard report
Math Word Problem Solving MATH DeepSeekMATH-RL-7B (greedy decoding) Accuracy 51.7 #44 of 135 Archive leaderboard report
Math Word Problem Solving MATH DeepSeekMATH-RL-7B (greedy decoding) Parameters (Billions) 7 #44 of 135 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

Absolute Position EncodingsAdamAttentionBPEDense ConnectionsDropoutEntropy RegularizationGPT-4Label SmoothingLayer NormalizationLinear LayerMulti-Head AttentionPPOPosition-Wise Feed-Forward LayerResidual ConnectionSoftmaxTransformer

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