SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2604.08281 · 2026

When to Trust Tools? Adaptive Tool Trust Calibration For Tool-Integrated Math Reasoning

Dong Li, Juntao Li, Min Zhang, Yu Luo, Yixin Ji, Jinpeng Li, Peifeng Li, Ruotao Xu

arXiv · PDF · Open in the Atlas

Code that ran

7 of 8 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; 1 honoured the contract we drafted, 1 violated it and 0 have no contract verdict; 5 failures of our instrument; 1 no recorded run), measured 2026-11-22; 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/2604.08281.json · badge: /api/ran/2604.08281.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
00Dreamer00/ATTC canonical (classifier label)
classifier basis: census
7 ran · 1 no recorded run
FunctionStatusWhere it lives
calculate_code_prob_avg Ran
ran; our contract draft was wrong, not the code
00Dreamer00/ATTC/math_eval_confident.py
code served (permissive licence) · get_code("def39042747f870e")
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 float64, ndim 0, 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 math
from scipy.stats import gmean

class SimpleTokenizer:
    def encode(self, text, add_special_tokens=False):
        # Simple tokenization: split by spaces and punctuation
        words = text.split()
        return [hash(word) % 10000 for word in words]
    
    def decode(self, token_ids):
        # Simple decode: return a fixed-length string per token
        return "tok"

output_text = "Here is some code:\n```python\ndef hello():\n    return 42\n```\nEnd of code."

tokenizer = SimpleTokenizer()

# Get tokens for the output text
tokens = tokenizer.encode(output_text, add_special_tokens=False)
num_tokens = len(tokens)

# Create log_probs as a list of reasonable log probabilities
log_probs = [-0.5 - 0.1 * i for i in range(num_tokens)]

ARGS = {
    "output_text": output_text,
    "log_probs": log_probs,
    "tokenizer": tokenizer
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
scipy — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type float64, ndim 0, dtype family float — no output shape is stated: observe_signature.py recorded float ndim 0, which disagrees with the fixture run
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
extract_multi_choice_answer Ran
ran; our contract draft was wrong, not the code
00Dreamer00/ATTC/parser.py
pointer only (licence: MIT); the stored copy get_code serves, from QwenLM/Qwen2.5-Math, is pointer only (licence: NONE) · get_code("ed2f4689128f9885")
How it ran

The body of this sample is withheld: this paper's link allows the body inline (MIT), but get_code serves the stored copy, first harvested from QwenLM/Qwen2.5-Math, and GitHub detected no licence in that repository when it was checked. The setup block that called it is withheld with it, because it is written against that body. source_path and origin say where the code lives, to read under the repository's own terms. A run is recorded.

generate_completions Ran
ran; our contract draft was wrong, not the code
00Dreamer00/ATTC/model_utils.py
pointer only (licence: MIT); the stored copy get_code serves, from fanqiwan/FuseAI, is pointer only (licence: NONE) · get_code("1525d8bc0fc8d5a0")
How it ran

The body of this sample is withheld: this paper's link allows the body inline (MIT), but get_code serves the stored copy, first harvested from fanqiwan/FuseAI, and GitHub detected no licence in that repository when it was checked. The setup block that called it is withheld with it, because it is written against that body. source_path and origin say where the code lives, to read under the repository's own terms. A run is recorded.

is_digit Ran
ran, violated the contract we drafted
00Dreamer00/ATTC/grader.py
pointer only (licence: MIT); the stored copy get_code serves, from QwenLM/Qwen2.5-Math, is pointer only (licence: NONE) · get_code("fb37d3c84b18e368")
How it ran

The body of this sample is withheld: this paper's link allows the body inline (MIT), but get_code serves the stored copy, first harvested from QwenLM/Qwen2.5-Math, and GitHub detected no licence in that repository when it was checked. The setup block that called it is withheld with it, because it is written against that body. source_path and origin say where the code lives, to read under the repository's own terms. A run is recorded.

parse_digits Ran
ran, honoured the contract we drafted
00Dreamer00/ATTC/grader.py
pointer only (licence: MIT); the stored copy get_code serves, from QwenLM/Qwen2.5-Math, is pointer only (licence: NONE) · get_code("7d4b7f8c3875b126")
How it ran

The body of this sample is withheld: this paper's link allows the body inline (MIT), but get_code serves the stored copy, first harvested from QwenLM/Qwen2.5-Math, and GitHub detected no licence in that repository when it was checked. The setup block that called it is withheld with it, because it is written against that body. source_path and origin say where the code lives, to read under the repository's own terms. A run is recorded.

process_question Ran
ran; our contract draft was wrong, not the code
00Dreamer00/ATTC/add_new_dataset.py
code served (permissive licence) · get_code("48fad3d8f7e025e9")
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 str. 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 torch

def process_question(question):
    if '\n\n' in question:
        return question.split('\n\n', 1)[1]
    return question

ARGS = {
    'question': 'What is this?\n\nThis is the answer.'
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type str — a str output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a str 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
str_to_pmatrix Ran
ran; our contract draft was wrong, not the code
00Dreamer00/ATTC/grader.py
code served (permissive licence) · get_code("ad19a7a80772794c")
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 str. 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 re

def str_to_pmatrix(input_str):
    input_str = input_str.strip()
    matrix_str = re.findall('\\{.*,.*\\}', input_str)
    pmatrix_list = []
    for m in matrix_str:
        m = m.strip('{}')
        pmatrix = '\\begin{pmatrix}' + m.replace(',', '\\') + '\\end{pmatrix}'
        pmatrix_list.append(pmatrix)
    return ', '.join(pmatrix_list)


# Setup code to construct valid arguments
input_str_example = "{1,2,3,4} and {5,6,7,8}"

ARGS = {
    'input_str': input_str_example
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
none outside the standard library — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type str — a str output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a str 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
load_hf_lm_and_tokenizer Not yet run
unverified (harvested, no recorded run)
00Dreamer00/ATTC/model_utils.py
pointer only (licence: MIT); the stored copy get_code serves, from YangLing0818/SuperCorrect-llm, is pointer only (licence: NONE) · get_code("ee71a94d9dd1b8e9")
How it ran

The body of this sample is withheld: this paper's link allows the body inline (MIT), but get_code serves the stored copy, first harvested from YangLing0818/SuperCorrect-llm, and GitHub detected no licence in that repository when it was checked. The setup block that called it is withheld with it, because it is written against that body. source_path and origin say where the code lives, to read under the repository's own terms. No invocation recorded.

Also present in other papers' code

Of 8 samples lifted from this paper's linked repositories, 8 carry a function identity and 0 do not. Of the 8 distinct functions with an identity, 6 are also present in other papers whose code we hold (6 with at least one paper dated earlier, 5 the same month, 2 later) and 2 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.

FunctionAlso present inEarliest-datedOthers named
is_digit 18 other papers
16 dated earlier, 1 the same month, 1 undated
2405.14365 (2024-05)
JiuZhang3.0: Efficiently Improving Mathematical Reasoning by Training Small Data Synthesis Models
2410.07985 (2024-10)
Omni-MATH: A Universal Olympiad Level Mathematic Benchmark For Large Language Models
2502.06703 (2025-02)
Can 1B LLM Surpass 405B LLM? Rethinking Compute-Optimal Test-Time Scaling
and 15 more not named here
load_hf_lm_and_tokenizer 18 other papers
15 dated earlier, 2 the same month, 1 later
2405.14365 (2024-05)
JiuZhang3.0: Efficiently Improving Mathematical Reasoning by Training Small Data Synthesis Models
2408.07471 (2024-08)
Bridging and Modeling Correlations in Pairwise Data for Direct Preference Optimization
2410.07985 (2024-10)
Omni-MATH: A Universal Olympiad Level Mathematic Benchmark For Large Language Models
and 15 more not named here
parse_digits 18 other papers
16 dated earlier, 1 the same month, 1 undated
2405.14365 (2024-05)
JiuZhang3.0: Efficiently Improving Mathematical Reasoning by Training Small Data Synthesis Models
2410.07985 (2024-10)
Omni-MATH: A Universal Olympiad Level Mathematic Benchmark For Large Language Models
2502.06703 (2025-02)
Can 1B LLM Surpass 405B LLM? Rethinking Compute-Optimal Test-Time Scaling
and 15 more not named here
generate_completions 17 other papers
14 dated earlier, 2 the same month, 1 later
2405.14365 (2024-05)
JiuZhang3.0: Efficiently Improving Mathematical Reasoning by Training Small Data Synthesis Models
2408.07471 (2024-08)
Bridging and Modeling Correlations in Pairwise Data for Direct Preference Optimization
2410.07985 (2024-10)
Omni-MATH: A Universal Olympiad Level Mathematic Benchmark For Large Language Models
and 14 more not named here
extract_multi_choice_answer 12 other papers
10 dated earlier, 1 the same month, 1 undated
2405.14365 (2024-05)
JiuZhang3.0: Efficiently Improving Mathematical Reasoning by Training Small Data Synthesis Models
2408.03314 (2024-08)
Scaling LLM Test-Time Compute Optimally can be More Effective than Scaling Model Parameters
2410.07985 (2024-10)
Omni-MATH: A Universal Olympiad Level Mathematic Benchmark For Large Language Models
and 9 more not named here
str_to_pmatrix 1 other paper
1 dated earlier
2405.14365 (2024-05)
JiuZhang3.0: Efficiently Improving Mathematical Reasoning by Training Small Data Synthesis Models

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 reasoning models (LRMs) have achieved strong performance enhancement through scaling test time computation, but due to the inherent limitations of the underlying language models, they still have shortcomings in tasks that require precise computation and extensive knowledge reserves. Tool-Integrated Reasoning (TIR) has emerged as a promising paradigm that incorporates tool call and execution within the reasoning trajectory. Although recent works have released some powerful open-source TIR models, our analysis reveals that these models still suffer from critical deficiencies. We find that when the reasoning of the model conflicts with the tool results, the model tends to believe in its own reasoning. And there are cases where the tool results are correct but are ignored by the model, resulting in incorrect answers, which we define as "Tool Ignored". This indicates that the model does not know when to trust or ignore the tool. To overcome these limitations, We introduce Adaptive Tool Trust Calibration (ATTC), a novel framework that guides the model to adaptively choose to trust or ignore the tool results based on the confidence score of generated code blocks. The experimental results from various open-source TIR models of different sizes and across multiple datasets demonstrate that ATTC effectively reduces the "Tool Ignored" issue, resulting in a performance increase of 4.1% to 7.5%. Our codes are available at https://github.com/00Dreamer00/ATTC.

For agents

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

get_harvested_code_for_paper("2604.08281")
get_code_for_paper("2604.08281")
have("2604.08281")

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

curl https://syntology.ai/api/ran/2604.08281.json

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

[![Syntology run record](https://syntology.ai/api/ran/2604.08281.svg)](https://syntology.ai/paper/2604.08281)

Connect an agent — have() is free.