SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2211.09085 · 2022

Galactica: A Large Language Model for Science

arXiv · PDF · Open in the Atlas

Code that ran

2 of 2 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; 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/2211.09085.json · badge: /api/ran/2211.09085.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
paperswithcode/galai canonical (classifier label)
classifier basis not recorded on the link
2 ran · 0 no recorded run
FunctionStatusWhere it lives
escape_custom_split_sequence Ran
ran; our contract draft was wrong, not the code
paperswithcode/galai/galai/utils.py
code served (permissive licence) · get_code("779d2e3138d09f2d")
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

CUSTOM_SEQ_RE = re.compile('(\\[START_(DNA|SMILES|I_SMILES|AMINO)])(.*?)(\\[END_\\2])')

SPLIT_MARKER = f'SPL{1}T-TH{1}S-Pl3A5E'

def _insert_split_marker(m: re.Match):
    """
    Applies split marker based on a regex match of special tokens such as
    [START_DNA].

    Parameters
    ----------
    n : str
        Input text to split

    Returns
    ----------
    str - the text with the split token added
    """
    start_token, _, sequence, end_token = m.groups()
    sequence = re.sub('(.)', f'{SPLIT_MARKER}\\1', sequence, flags=re.DOTALL)
    return f'{start_token}{sequence}{SPLIT_MARKER}{end_token}'

def escape_custom_split_sequence(text):
    """
    Applies custom splitting to the text for GALILEO's tokenization

    Parameters
    ----------
    text : str
        Input text to split

    Returns
    ----------
    str - the text with the split token added
    """
    return CUSTOM_SEQ_RE.sub(_insert_split_marker, text)


# Setup code to create valid arguments
ARGS = {
    "text": "[START_DNA]ACGTACGT[END_DNA] some other text [START_SMILES]CCO[END_SMILES]"
}
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
extract_references_from_text Ran
ran on a synthesized input
paperswithcode/galai/galai/utils.py
code served (permissive licence) · get_code("282223027b104a81")
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 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 re
from typing import List

REFERENCE_RE = re.compile('\\[START_REF\\](.*?)\\[END_REF\\]', flags=re.DOTALL)

def extract_references_from_text(text: str) -> List[str]:
    return [cit.strip() for cit in REFERENCE_RE.findall(text)]

# Construct test text with reference markers
text_with_references = """
This is some introduction text.
[START_REF]
Author, A. (2020). Paper Title. Journal Name, 10(2), 123-145.
[END_REF]
More content here.
[START_REF]
  Smith, B. & Jones, C. (2021). Another Paper. Conference Proceedings.  
[END_REF]
Final conclusion.
"""

ARGS = {
    "text": text_with_references
}
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 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

Also present in other papers' code

Of 2 samples lifted from this paper's linked repositories, 2 carry a function identity and 0 do not. Of the 2 distinct functions with an identity, 1 are also present in other papers whose code we hold (0 with at least one paper dated earlier, 0 the same month, 1 later) and 1 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
escape_custom_split_sequence 4 other papers
4 later
2310.12798 (2023-10)
MolCA: Molecular Graph-Language Modeling with Cross-Modal Projector and Uni-Modal Adapter
2024.findings-acl.318 (2024 (year only))
ReactXT: Understanding Molecular "Reaction-ship" via Reaction-Contextualized Molecule-Text Pretraining
2401.13923 (2024-01)
Towards 3D Molecule-Text Interpretation in Language Models
and 1 more not named here

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

Information overload is a major obstacle to scientific progress. The explosive growth in scientific literature and data has made it ever harder to discover useful insights in a large mass of information. Today scientific knowledge is accessed through search engines, but they are unable to organize scientific knowledge alone. In this paper we introduce Galactica: a large language model that can store, combine and reason about scientific knowledge. We train on a large scientific corpus of papers, reference material, knowledge bases and many other sources. We outperform existing models on a range of scientific tasks. On technical knowledge probes such as LaTeX equations, Galactica outperforms the latest GPT-3 by 68.2% versus 49.0%. Galactica also performs well on reasoning, outperforming Chinchilla on mathematical MMLU by 41.3% to 35.7%, and PaLM 540B on MATH with a score of 20.4% versus 8.8%. It also sets a new state-of-the-art on downstream tasks such as PubMedQA and MedMCQA dev of 77.6% and 52.9%. And despite not being trained on a general corpus, Galactica outperforms BLOOM and OPT-175B on BIG-bench. We believe these results demonstrate the potential for language models as a new interface for science. We open source the model for the benefit of the scientific community.

For agents

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

get_harvested_code_for_paper("2211.09085")
get_code_for_paper("2211.09085")
have("2211.09085")

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

curl https://syntology.ai/api/ran/2211.09085.json

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

[![Syntology run record](https://syntology.ai/api/ran/2211.09085.svg)](https://syntology.ai/paper/2211.09085)

Connect an agent — have() is free.