Papers › Math Word Problem Solving by Generating Linguistic Variants of Problem Statements

Math Word Problem Solving by Generating Linguistic Variants of Problem Statements

24 Jun 2023arXiv:2306.13899archive 2025-07-28

Syed Rifat Raiyan, Md. Nafis Faiyaz, Shah Md. Jawad Kabir, Mohsinul Kabir, Hasan Mahmud, Md Kamrul Hasan

The art of mathematical reasoning stands as a fundamental pillar of intellectual progress and is a central catalyst in cultivating human ingenuity. Researchers have recently published a plethora of works centered around the task of solving Math Word Problems (MWP) - a crucial stride towards general AI. These existing models are susceptible to dependency on shallow heuristics and spurious correlations to derive the solution expressions. In order to ameliorate this issue, in this paper, we propose a framework for MWP solvers based on the generation of linguistic variants of the problem text. The approach involves solving each of the variant problems and electing the predicted expression with the majority of the votes. We use DeBERTa (Decoding-enhanced BERT with disentangled attention) as the encoder to leverage its rich textual representations and enhanced mask decoder to construct the solution expressions. Furthermore, we introduce a challenging dataset, PARAMAWPS, consisting of paraphrased, adversarial, and inverse variants of selectively sampled MWPs from the benchmark MAWPS dataset. We extensively experiment on this dataset along with other benchmark datasets using some baseline MWP solver models. We show that training on linguistic variants of problem statements and voting on candidate predictions improve the mathematical reasoning and robustness of the model. We make our code and data publicly available.

PaperPDFCode

Code

starscream-11813/variational-mathematical-reasoning officialmentioned in papermentioned on GitHub 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

Not run by Syntology. Nothing on this page verifies that the listed code works.

Tasks

DecoderIngenuityMathMath Word Problem SolvingMathematical Reasoning

Datasets

Introduced by this paper, per the archive.

ParaMAWPS

Results from the paper archive 2025-07-28

TaskDatasetModelMetricValueRank at snapshotLeaderboardReport
Math Word Problem Solving MAWPS DeBERTa (PM + VM) Accuracy (%) 91.0 #8 of 25 Archive leaderboard report
Math Word Problem Solving MAWPS GPT-3.5 turbo (175B) Accuracy (%) 80.3 #17 of 25 Archive leaderboard report
Math Word Problem Solving MAWPS GPT-J Accuracy (%) 9.9 #21 of 25 Archive leaderboard report
Math Word Problem Solving MAWPS GPT-3 text-curie-001 (13B) Accuracy (%) 4.09 #24 of 25 Archive leaderboard report
Math Word Problem Solving MAWPS GPT-3 text-babbage-001 (6.7B) Accuracy (%) 2.76 #25 of 25 Archive leaderboard report
Math Word Problem Solving ParaMAWPS DeBERTa (VM) Accuracy (%) 79.1 #1 of 6 Archive leaderboard report
Math Word Problem Solving ParaMAWPS GPT-3.5 Turbo (175B) Accuracy (%) 73.0 #3 of 6 Archive leaderboard report
Math Word Problem Solving ParaMAWPS GPT-J (6B) Accuracy (%) 5.9 #4 of 6 Archive leaderboard report
Math Word Problem Solving ParaMAWPS GPT-3 text-curie-001 (13B) Accuracy (%) 4.20 #5 of 6 Archive leaderboard report
Math Word Problem Solving ParaMAWPS GPT-3 text-babbage-001 (6.7B) Accuracy (%) 3.21 #6 of 6 Archive leaderboard report
Math Word Problem Solving SVAMP DeBERTa Accuracy 63.5 #13 of 26 Archive leaderboard report
Math Word Problem Solving SVAMP DeBERTa Execution Accuracy 63.5 #13 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

AdamAttentionAttention DropoutBERTDeBERTaDense ConnectionsDropoutLayer NormalizationLinear LayerLinear Warmup With Linear DecayMulti-Head AttentionResidual ConnectionSoftmaxWeight DecayWordPiece

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