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 ran

In Syntology Open this paper in Syntology's Atlas, the map of the papers in Syntology's graph and their citations.

For agents, Syntology's MCP tool lists every function and class Syntology harvested from this paper and whether it ran (how to connect): get_harvested_code_for_paper(arxiv_id="2402.03300")

Code

Syntology Ran 8 of 24 code samples harvested from 3 repositories linked to this paper; 16 have no recorded run. Of those that ran: 2 ran · our draft was wrong; 2 ran · fixture could not drive it; 4 ran with no contract checked.

By repository: official repository: 9 samples from 1 repository, 4 ran; community (archive-listed): 15 samples from 2 repositories, 4 ran. The run record, sample by sample. “Ran” means executed on a synthesized input, not that the code is correct or reproduces the paper.

deepseek-ai/deepseek-math officialmentioned in papermentioned on GitHubpytorchMIT report
openpsi-project/realhf mentioned on GitHubpytorch report
shibing624/medicalgpt mentioned on GitHubpytorchApache-2.0 report
synthlabsai/big-math mentioned on GitHubMIT report
yynil/rwkvinside mentioned on GitHubpytorchApache-2.0 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

24 samples harvested; 8 ran; 0 honoured the contract we drafted; 16 have no recorded run. Read from Syntology's graph 2026-09-24; that is when this build read the record, not when the samples ran.

2ran · our draft was wrong
2ran · fixture could not drive it
4ran
16unverified

Licence: 0 of the 24 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 3 repositories linked to this paper, official or community; each sample names its own and says which. “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. “Honoured” and “violated” refer to a contract Syntology drafted from the code itself; “our draft was wrong” and “fixture could not drive it” are failures of Syntology's instrument, not of the code.

Each sample ends with its code_sha256, Syntology's identity for that exact code. An agent fetches the stored sample with Syntology's MCP tool get_code(code_sha256="…") (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 from a repository marked official are listed first. 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 four 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; or the licence is outside the permissive list Syntology serves text under (MIT, Apache-2.0, BSD and similar). 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.

extract_boxed_answers deepseek-ai/deepseek-math/evaluation/data_processing/answer_extraction.py official repository ran fingerprinted MIT (permissive) · 6390dea763ce122e · report
extract_program_output deepseek-ai/deepseek-math/evaluation/data_processing/answer_extraction.py official repository ran fingerprinted MIT (permissive) · 0772e5020f6b0849 · report
shuffle deepseek-ai/deepseek-math/evaluation/utils.py official repository ran fingerprinted MIT (permissive) · 1d4da55afb0377ec · report
strip_string deepseek-ai/deepseek-math/evaluation/data_processing/answer_extraction.py official repository ran fingerprinted MIT (permissive) · 5abed389274c4c03 · report
aggregate_metrics deepseek-ai/deepseek-math/evaluation/summarize_results.py official repository unverified MIT (permissive) · 8b9e2baf73bb9a6f · report
aggregate_predictions deepseek-ai/deepseek-math/evaluation/summarize_results.py official repository unverified MIT (permissive) · 33425041e7f88b7f · report
check_proof deepseek-ai/deepseek-math/evaluation/unsafe_score_minif2f_isabelle.py official repository unverified MIT (permissive) · c59c51644b25a189 · report
markup_question deepseek-ai/deepseek-math/evaluation/run_subset_parallel.py official repository unverified MIT (permissive) · de0231790ab22287 · report
read_data deepseek-ai/deepseek-math/evaluation/utils.py official repository unverified MIT (permissive) · 3deb39de933b985a · report
repeat_kv yynil/rwkvinside/rwkv_inside/modeling_arwkv.py community (archive-listed) ran · fixture could not drive it fingerprinted Apache-2.0 (permissive) · 30d7eec482ebf6b1 · report
actor_loss_fn openpsi-project/realhf/realhf/impl/model/utils/ppo_functional.py community (archive-listed) ran · fixture could not drive it Apache-2.0 (permissive) · e2e940c65c7460d3 · report
apply_rotary_pos_emb yynil/rwkvinside/rwkv_inside/modeling_arwkv.py community (archive-listed) ran · our draft was wrong Apache-2.0 (permissive) · bac65c3dafaec040 · report
rotate_half yynil/rwkvinside/rwkv_inside/modeling_arwkv.py community (archive-listed) ran · our draft was wrong fingerprinted Apache-2.0 (permissive) · b99eea6376d1e212 · report
calculate_language_consistency_score yynil/rwkvinside/rl/deepspeed_rl_train.py community (archive-listed) unverified Apache-2.0 (permissive) · 8ddbdfd0041abedc · report
calculate_param_sizes yynil/rwkvinside/train_scripts/split_model_params.py community (archive-listed) unverified Apache-2.0 (permissive) · 93678d5fdf25571d · report
calculate_tensor_size yynil/rwkvinside/train_scripts/split_model_params.py community (archive-listed) unverified Apache-2.0 (permissive) · bc474830fb3ccd5d · report
compare_latex_numbers yynil/rwkvinside/rl/utilities.py community (archive-listed) unverified Apache-2.0 (permissive) · 620dc092b02e33f7 · report
distribute_params yynil/rwkvinside/train_scripts/split_model_params.py community (archive-listed) unverified Apache-2.0 (permissive) · d0a5e652d9881e59 · report
grpo_loss_with_old_logps yynil/rwkvinside/rl/grpo_trainer.py community (archive-listed) unverified Apache-2.0 (permissive) · f9c058ca4549e9f6 · report
parse_coordinate yynil/rwkvinside/rl/utilities.py community (archive-listed) unverified Apache-2.0 (permissive) · eb7f513a0e05ec8b · report
parse_latex_to_python yynil/rwkvinside/rl/utilities.py community (archive-listed) unverified Apache-2.0 (permissive) · f7cc3a0c10540f2e · report
reward_function yynil/rwkvinside/rl/train_rl_model.py community (archive-listed) unverified Apache-2.0 (permissive) · f872aa95062d3ac2 · report
selective_log_softmax_old yynil/rwkvinside/rl/grpo_trainer.py community (archive-listed) unverified Apache-2.0 (permissive) · 7c05d945015e9e96 · report
time_function yynil/rwkvinside/rl/profiler.py community (archive-listed) unverified Apache-2.0 (permissive) · b016e9b55d3c8dad · report

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