SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2202.13169 · ICLR · 2022

A systematic evaluation of large language models of code

Graham Neubig, Uri Alon, Frank Xu, Vincent Hellendoorn

arXiv · PDF · Open in the Atlas

Code that ran

2 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, 1 violated it and 0 have no contract verdict; 1 failure of our instrument; 1 no recorded run); 1 of them was also run on the host, not in the jail, by the peer-agreement lane that set its level, 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/2202.13169.json · badge: /api/ran/2202.13169.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
vhellendoorn/code-lms canonical (classifier label)
classifier basis: classifier_2025
2 ran · 1 no recorded run
FunctionStatusWhere it lives
ppl Ran
ran, violated the contract we drafted
vhellendoorn/code-lms/Evaluation/eval_codex_all.py
code served (permissive licence) · get_code("345c4c7d073c29d9")
How it ran

Ran in Syntology's jail on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1) and returned float, ndim 0, dtype family float, shape () as a later run recorded it. The image it ran in was not recorded. The comparison behind this sample's level is a different run: the peer-agreement lane ran it again on the host, not in the jail (report dated 2026-09-04T23:13:13Z). 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 math
import torch

torch.manual_seed(0)

# The function ppl takes avg_logprob as a parameter
# avg_logprob should be a numeric value (float or tensor that can be converted to float)
# The function computes 2 ** (-avg_logprob / math.log(2))

# Create a small float value for avg_logprob
avg_logprob_value = -2.5

ARGS = {
    "avg_logprob": avg_logprob_value
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1
Imports
torch — 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
get_hf_state_dict_from_pt_files Ran
ran; our contract draft was wrong, not the code
vhellendoorn/code-lms/Convert2HF/convert_neox_pt_to_huggingface_neox.py
code served (permissive licence) · get_code("5b631374e3ae19f9")
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) and returned OrderedDict. 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
import torch
import tempfile
from collections import OrderedDict

torch.manual_seed(0)

# Create a temporary directory with mock layer files
temp_dir = tempfile.mkdtemp()

# Create mock layer files with state dict content
layer_0_state = OrderedDict([
    ('word_embeddings.weight', torch.randn(4, 3)),
    ('attention.self.query.weight', torch.randn(4, 4)),
    ('_layernorm.weight', torch.randn(4)),
])

layer_1_state = OrderedDict([
    ('mlp.dense_h_to_4h.weight', torch.randn(5, 4)),
    ('_layernorm.bias', torch.randn(4)),
])

norm_state = OrderedDict([
    ('norm.weight', torch.randn(4)),
])

final_state = OrderedDict([
    ('final_linear.weight', torch.randn(6, 4)),
])

# Save mock layer files
torch.save(layer_0_state, os.path.join(temp_dir, 'layer_0.pt'))
torch.save(layer_1_state, os.path.join(temp_dir, 'layer_1.pt'))
torch.save(norm_state, os.path.join(temp_dir, 'norm.pt'))
torch.save(final_state, os.path.join(temp_dir, 'final.pt'))

ARGS = {
    'checkpoint_dir': temp_dir
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1
Imports
torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type OrderedDict — a OrderedDict output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a OrderedDict 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_model_and_generate Not yet run
unverified (harvested, no recorded run)
vhellendoorn/code-lms/Convert2HF/generate.py
code served (permissive licence) · get_code("6bdf1b1bc002986d")
How it ran

no invocation recorded.

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

Large language models (LMs) of code have recently shown tremendous promise in completing code and synthesizing code from natural language descriptions. However, the current state-of-the-art code LMs (e.g., Codex (Chen et al., 2021)) are not publicly available, leaving many questions about their model and data design decisions. We aim to fill in some of these blanks through a systematic evaluation of the largest existing models: Codex, GPT-J, GPT-Neo, GPT-NeoX-20B, and CodeParrot, across various programming languages. Although Codex itself is not open-source, we find that existing open-source models do achieve close results in some programming languages, although targeted mainly for natural language modeling. We further identify an important missing piece in the form of a large open-source model trained exclusively on a multi-lingual corpus of code. We release a new model, PolyCoder, with 2.7B parameters based on the GPT-2 architecture, that was trained on 249GB of code across 12 programming languages on a single machine. In the C programming language, PolyCoder outperforms all models including Codex. Our trained models are open-source and publicly available at https://github.com/VHellendoorn/Code-LMs, which enables future research and application in this area.

For agents

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

get_harvested_code_for_paper("2202.13169")
get_code_for_paper("2202.13169")
have("2202.13169")

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

curl https://syntology.ai/api/ran/2202.13169.json

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

[![Syntology run record](https://syntology.ai/api/ran/2202.13169.svg)](https://syntology.ai/paper/2202.13169)

Connect an agent — have() is free.