SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2303.05398 · ACL · 2023

MathPrompter: Mathematical Reasoning using Large Language Models

Liang Du, Harsh Shrivastava, Shima Imani

arXiv · PDF · Open in the Atlas

Code that ran

2 of 5 samples (functions and classes) lifted from this paper's linked repositories executed on a synthesized input in Syntology's jail (0 of them constructed an object rather than computing a result; 0 honoured the contract we drafted, 0 violated it and 1 has no contract verdict; 1 failure of our instrument; 3 no recorded run), measured 2026-11-22; repository links from the Papers with Code archive attributed under CC BY-SA, no affiliation; bodies inline only under a permissive licence as GitHub detects it. Not a reproduction of the paper's results.

"Ran" means the function executed on a synthesized input and returned a value. It is not a reproduction of the paper's results. This record, dated, as JSON: /api/ran/2303.05398.json · badge: /api/ran/2303.05398.svg

Each repository's role is a label: a classifier's output recorded on the link, or, where no role is recorded, what the Papers with Code archive marked. The line under it names its basis. Neither says who wrote the code.

RepositoryRoleRan
RamonKaspar/MathPrompter community
Papers with Code archive
2 ran · 3 no recorded run
FunctionStatusWhere it lives
OpenAIService Ran
ran on a synthesized input
RamonKaspar/MathPrompter/math_prompts_generation.py
code served (permissive licence) · get_code("da81018d427bc031")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, class construct-then-call prompt, model not recorded) and returned list, ndim 1, dtype family float; no output shape is stated. The image it ran in was not recorded. If your copy returns the same type and ndim on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import os
from unittest.mock import Mock, patch

# Mock the OpenAI client to avoid actual API calls
mock_client = Mock()
mock_response = Mock()
mock_response.choices = [Mock()]
mock_response.choices[0].message.content = "This is a generated response from the language model."
mock_client.chat.completions.create.return_value = mock_response

# Patch the OpenAI constructor to return our mock
with patch('openai.OpenAI', return_value=mock_client):
    # Set a dummy API key in environment
    os.environ['OPENAI_API_KEY'] = 'dummy-key-for-testing'
    
    INIT_ARGS = {
        'model_name': 'gpt-3.5-turbo',
        'temperature': 0.7,
        'max_tokens': 150
    }
    
    ARGS = {
        'system_prompt': 'You are a helpful assistant.',
        'few_shot_examples': [
            {'question': 'What is 2+2?', 'answer': '4'},
            {'question': 'What is the capital of France?', 'answer': 'Paris'}
        ],
        'question': 'What is the meaning of life?'
    }
    
    METHOD = 'create_prompt'
Setup block drafted by
a language model, class construct-then-call prompt, model not recorded
Imports
openai — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type list, ndim 1, dtype family float — a list output has no array shape of its own; none is stated
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
get_llm_service Ran
ran; our contract draft was wrong, not the code
RamonKaspar/MathPrompter/math_prompts_generation.py
code served (permissive licence) · get_code("bf1b6f89bf903fbc")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned OpenAIService. The image it ran in was not recorded. If your copy returns the same type on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import os

# Set up environment variables for API keys (dummy values since we won't make real requests)
os.environ['OPENAI_API_KEY'] = 'sk-test-key-dummy'
os.environ['AZURE_OPENAI_KEY'] = 'azure-test-key-dummy'
os.environ['AZURE_OPENAI_ENDPOINT'] = 'https://test.openai.azure.com/'

# Define arguments for get_llm_service function
ARGS = {
    'service_name': 'openai',
    'model_name': 'gpt-3.5-turbo',
    'temperature': 0.7,
    'max_tokens': 256
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
openai — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type OpenAIService — a OpenAIService output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a OpenAIService output the recorded ndim 0 and dtype family float describe that coercion, not the output, and are not stated
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
AzureOpenAIService Not yet run
unverified (harvested, no recorded run)
RamonKaspar/MathPrompter/math_prompts_generation.py
code served (permissive licence) · get_code("905603ae58189fc7")
How it ran

no invocation recorded.

LLMInterface Not yet run
unverified (harvested, no recorded run)
RamonKaspar/MathPrompter/math_prompts_generation.py
code served (permissive licence) · get_code("06c040ff02496ef2")
How it ran

no invocation recorded.

algebric_expression_generation Not yet run
unverified (harvested, no recorded run)
RamonKaspar/MathPrompter/math_prompts_generation.py
code served (permissive licence) · get_code("907831f47445bd3f")
How it ran

no invocation recorded.

Also present in other papers' code

Of 5 samples lifted from this paper's linked repositories, 5 carry a function identity and 0 do not. Of the 5 distinct functions with an identity, 0 are also present in other papers whose code we hold (0 with at least one paper dated earlier, 0 the same month, 0 later) and 5 are only here among the 52,977 papers whose code Syntology holds (as of 2026-09-26).

Only here means: among the 52,977 papers whose code Syntology holds (51,887 with lifted samples, 1,090 through their assemblies only), as of 2026-09-26, no other paper's lifted samples or current assembly bricks contain an identical function body after normalisation.

Presence, not direction: an identical function body in two papers' repositories does not say who wrote it, which came from which, or that either paper's method rests on it. Matching is exact after normalisation, so a renamed or edited function is not matched. Dates are when papers appeared, not when code was written. Only .py files the harvester lifted are covered; notebooks are not read. A new-style arXiv id YYMM.NNNN dates a paper to the month 20YY-MM; otherwise Paper.year dates it to the year; otherwise it is undated. Dates are when the paper appeared, not when the code was written. Read 2026-11-22.

Repositories linked to this paper

Some links come from the archived Papers with Code dataset (CC BY-SA 4.0): attribution and licence.

Abstract

Large Language Models (LLMs) have limited performance when solving arithmetic reasoning tasks and often provide incorrect answers. Unlike natural language understanding, math problems typically have a single correct answer, making the task of generating accurate solutions more challenging for LLMs. To the best of our knowledge, we are not aware of any LLMs that indicate their level of confidence in their responses which fuels a trust deficit in these models impeding their adoption. To address this deficiency, we propose 'MathPrompter', a technique that improves performance of LLMs on arithmetic problems along with increased reliance in the predictions. MathPrompter uses the Zero-shot chainof-thought prompting technique to generate multiple Algebraic expressions or Python functions to solve the same math problem in different ways and thereby raise the confidence level in the output results. This is in contrast to other prompt based CoT methods, where there is no check on the validity of the intermediate steps followed. Our technique improves over state-of-the-art on the MultiArith dataset (78.7% → 92.5%) evaluated using 175B parameter GPT-based LLM.

For agents

The same record, over MCP at https://syntology.ai/mcp:

get_harvested_code_for_paper("2303.05398")
get_code_for_paper("2303.05398")
have("2303.05398")

The run record, dated, one paper per request, free:

curl https://syntology.ai/api/ran/2303.05398.json

A badge for a README (the split and the date, never a ratio):

[![Syntology run record](https://syntology.ai/api/ran/2303.05398.svg)](https://syntology.ai/paper/2303.05398)

Connect an agent — have() is free.