SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2304.07995 · 2023

From Zero to Hero: Examining the Power of Symbolic Tasks in Instruction Tuning

arXiv · PDF · Open in the Atlas

Code that ran

3 of 3 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 3 have no contract verdict; 0 failures of our instrument; 0 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/2304.07995.json · badge: /api/ran/2304.07995.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
sail-sg/symbolic-instruction-tuning canonical (classifier label)
classifier basis: census
3 ran · 0 no recorded run
FunctionStatusWhere it lives
evaluate_example Ran
ran on a synthesized input
sail-sg/symbolic-instruction-tuning/eval_utils.py
code served (permissive licence) · get_code("b6b4078e0666a60f")
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 bool, ndim 0, dtype family float, shape () as a later run recorded it. The image it ran in was not recorded. If your copy returns the same type, ndim and shape 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
import numpy as np
from collections import defaultdict

torch.manual_seed(0)

# The function takes two string parameters: predict_str and ground_str
# It compares them by splitting on commas and parsing as floats or strings
# The strings can contain numbers, text, or both separated by commas

ARGS = {
    "predict_str": "1.5, 2.0, hello",
    "ground_str": "1.5, 2.0, hello"
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy, torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type bool, ndim 0, dtype family float, shape () — shape () as a later run recorded it
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_denotation_accuracy Ran
ran on a synthesized input
sail-sg/symbolic-instruction-tuning/eval_utils.py
code served (permissive licence) · get_code("bf4f0ebba92d254d")
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 float, ndim 0, dtype family float, shape () as a later run recorded it. The image it ran in was not recorded. If your copy returns the same type, ndim and shape 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.

from collections import defaultdict
from typing import List

delimiter = ','

def evaluate_example(predict_str: str, ground_str: str):
    ground_str = ground_str.lower()
    predict_str = predict_str.lower()
    predict_spans = predict_str.split(delimiter)
    ground_spans = ground_str.split(delimiter)
    predict_values = defaultdict(lambda: 0)
    ground_values = defaultdict(lambda: 0)
    for span in predict_spans:
        try:
            predict_values[float(span)] += 1
        except ValueError:
            predict_values[span.strip()] += 1
    for span in ground_spans:
        try:
            ground_values[float(span)] += 1
        except ValueError:
            ground_values[span.strip()] += 1
    _is_correct = predict_values == ground_values
    return _is_correct

def get_denotation_accuracy(predictions: List[str], references: List[str], **kwargs):
    assert len(predictions) == len(references)
    correct_num = 0
    for predict_str, ground_str in zip(predictions, references):
        is_correct = evaluate_example(predict_str, ground_str)
        if is_correct:
            correct_num += 1
    return correct_num / len(predictions)

# Setup code
ARGS = {
    "predictions": [
        "1.5, 2.3, apple",
        "cat, dog, 3.14",
        "hello, world"
    ],
    "references": [
        "1.5, 2.3, apple",
        "CAT, DOG, 3.14",
        "HELLO, WORLD"
    ]
}
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 float, ndim 0, dtype family float, shape () — shape () as a later run recorded it
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
normalize_output Ran
ran on a synthesized input
sail-sg/symbolic-instruction-tuning/baseline/evaluate.py
code served (permissive licence) · get_code("94f609ea026fcf5c")
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
import torch.nn as nn

torch.manual_seed(0)

ARGS = {
    'output': 'The answer is hello world; this is great.'
}
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

Also present in other papers' code

Of 3 samples lifted from this paper's linked repositories, 3 carry a function identity and 0 do not. Of the 3 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 3 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

Fine-tuning language models on tasks with instructions has demonstrated potential in facilitating zero-shot generalization to unseen tasks. In this paper, we introduce a straightforward yet effective method for enhancing instruction tuning by employing symbolic tasks. Compared to crowdsourced human tasks or model-generated tasks, symbolic tasks present a unique advantage as they can be easily generated in vast quantities, theoretically providing an infinite supply of high-quality training instances. To explore the potential of symbolic tasks, we carry out an extensive case study on the representative symbolic task of SQL execution. Empirical results on various benchmarks validate that the integration of SQL execution leads to significant improvements in zero-shot scenarios, particularly in table reasoning. Notably, our 3B model surpasses both the 175B GPT-3 and ChatGPT in zero-shot table reasoning across four benchmarks. Furthermore, experimental results on BBH (27 tasks) and MMLU (57 tasks) reveal that language models can be enhanced through symbolic tasks without compromising their generality. We hope that our paper serves as a catalyst, inspiring increased efforts to incorporate symbolic tasks in instruction tuning.

For agents

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

get_harvested_code_for_paper("2304.07995")
get_code_for_paper("2304.07995")
have("2304.07995")

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

curl https://syntology.ai/api/ran/2304.07995.json

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

[![Syntology run record](https://syntology.ai/api/ran/2304.07995.svg)](https://syntology.ai/paper/2304.07995)

Connect an agent — have() is free.