Papers › Revisiting Grammatical Error Correction Evaluation and Beyond

Revisiting Grammatical Error Correction Evaluation and Beyond

3 Nov 2022arXiv:2211.01635archive 2025-07-28

Peiyuan Gong, Xuebo Liu, Heyan Huang, Min Zhang

Pretraining-based (PT-based) automatic evaluation metrics (e.g., BERTScore and BARTScore) have been widely used in several sentence generation tasks (e.g., machine translation and text summarization) due to their better correlation with human judgments over traditional overlap-based methods. Although PT-based methods have become the de facto standard for training grammatical error correction (GEC) systems, GEC evaluation still does not benefit from pretrained knowledge. This paper takes the first step towards understanding and improving GEC evaluation with pretraining. We first find that arbitrarily applying PT-based metrics to GEC evaluation brings unsatisfactory correlation results because of the excessive attention to inessential systems outputs (e.g., unchanged parts). To alleviate the limitation, we propose a novel GEC evaluation metric to achieve the best of both worlds, namely PT-M2 which only uses PT-based metrics to score those corrected parts. Experimental results on the CoNLL14 evaluation task show that PT-M2 significantly outperforms existing methods, achieving a new state-of-the-art result of 0.949 Pearson correlation. Further analysis reveals that PT-M2 is robust to evaluate competitive GEC systems. Source code and scripts are freely available at https://github.com/pygongnlp/PT-M2.

PaperPDFCodeCode Syntology ranMethod as code

In Syntology View this paper on Syntology, its page in Syntology's graph. That page adds which of its functions are also present in other papers' code Syntology holds, with dates, and its dated run record as JSON and a README badge. It also lists the repositories linked to the paper, the abstract and the calls for agents.

Open this paper in Syntology's Atlas, the map of the papers in Syntology's graph and their citations.

For agents, on Syntology's MCP service (how to connect):

Code

Syntology 31 code samples harvested from 1 repository linked to this paper: 21 ran (of which 0 constructed an object rather than computing a result; 6 with no instrument failure: 4 honoured, 2 violated, 0 with no contract checked; 15 where Syntology's instrument failed) · 10 unverified. Where Syntology's instrument failed: 15 where our contract draft was wrong; failures of Syntology's instrument, not of the code.

By repository: official repository: 31 samples from 1 repository, 21 ran (6 with no instrument failure; 15 where Syntology's instrument failed). The run record, sample by sample. “Ran” means executed on a synthesized input, not that the code is correct or reproduces the paper.

Levels written between 2026-09-22 and 2026-09-24 (UTC): the dates Syntology's graph wrote each sample's verification level, not when the code executed.

pygongnlp/pt-m2 officialmentioned in papermentioned on GitHubpytorchSyntology: 31 samples, 21 ran (15 instrument-failed) · 10 unverified report

Repository list and official/mentioned flags are the archive's, frozen 2025-07-28. Reachability, where shown, is from one Syntology probe window (2026-09-16 to 2026-09-18); repositories not probed show nothing. GitHub stars are not tracked.

Code Syntology ran Syntology

31 samples harvested: 21 ran (of which 0 constructed an object rather than computing a result; 6 with no instrument failure: 4 honoured, 2 violated, 0 with no contract checked; 15 where Syntology's instrument failed) · 10 unverified. “Where Syntology's instrument failed” counts failures of Syntology's instrument, not of the code. Syntology's record for this page has not changed since , the first build that kept a record date for it; when this build read Syntology's graph is in the build record. That date is not when the samples ran.

4ran · honoured contract
2ran · violated contract
15ran · our draft was wrong
10unverified

Licence: 31 of the 31 samples are pointer only, meaning Syntology does not serve that copy's text. This page shows no code text for any sample; each one links to its file in the repository.

Harvested from pygongnlp/pt-m2. “Ran” means the sample executed on a synthesized input. It does not mean the output is correct, and nothing here reproduces the paper's results. A line beginning “fixture:” under a sample that ran gives the arguments Syntology's fixture called the sample with, each as name:type/rank/element kind (? where none was recorded), and the type of what the fixture's run returned, exactly as the harness recorded them; the arguments and the returned type were recorded by two separate runs of the same fixture. No output value is shown, and a returned type says nothing about whether the output was right. “Level recorded” gives the UTC date when Syntology's graph wrote the sample's verification level, not when the code executed; where the level was written before Syntology's graph kept that date, the row says “level recorded, date not recorded”. “Honoured” and “violated” refer to a contract Syntology drafted from the code itself. “Our draft was wrong” and “the contract check could not run” are failures of Syntology's instrument, not of the code: the sample ran on Syntology's synthesized input, and then the drafted contract did not describe it, or the check itself raised an error (RAISES), lacked a dependency (DEP_MISSING) or ran out of time (TIMEOUT); hover the label for which. A metamorphic tier, where shown, is evidence of the output's shape, not that the sample implements the method.

Each sample ends with its code_sha256. A sample's code_sha256 is Syntology's identity for that exact code. It is printed in full (16 hexadecimal characters for every sample in this build), and get_code(code_sha256="…") takes it exactly as printed. An agent calls it on Syntology's MCP service (how to connect); click an identity to copy that call.

Repository labels, per sample. official repository: The archive marks this repository official for the paper. named in the paper: The archive records that the paper mentions this repository; it is not marked official. community (archive-listed): In the archive's code links for this paper, not marked official and not recorded as mentioned in the paper. found in paper text by Syntology: Syntology found this repository in the paper's own text; whether it is the authors' implementation is not asserted. community: Not in the archive's code links for this paper; a community repository Syntology harvested. Samples are grouped by repository: a repository marked official first, then by how directly the repository is tied to the paper, then in the archive's order; never by how many ran. Licence labels name the repository's licence as recorded at harvest. “Pointer only” means Syntology does not serve that copy's text, for one of six reasons: no licence file was found; the licence was not identified; the licence is recorded as permissive but that copy's record is not marked cleared; the licence is outside the permissive list Syntology serves text under (MIT, Apache-2.0, BSD and similar); this paper's copy was not located, so its licence is not recorded; or this paper's copy is cleared, but the stored copy of the same code that Syntology serves is not. Some licences outside that list permit redistribution, such as WTFPL, and GPL-3.0 under its conditions; they are simply not on the list. Hover a licence label for the reason. File links open the file on GitHub at the default branch, which may have changed since the harvest.

A function or class name links to its page when the name occurs in the harvested code of at least 5 papers; the name is a string, not an identity.

pygongnlp/pt-m2 official repository 31 samples, 21 ran (15 instrument-failed) · 10 unverified
best_edit_seq_bf metrics.py ran · our draft was wrong no licence file found · pointer only · 6685898ec6c99ed8 · report fixture: signature V:list/1/tuple, E:list/1/tuple, dist:dict/0/?, edits:dict/0/?, verby_verbose:bool/0/bool · returned list · level recorded 2026-09-24
comp_f1 metrics.py ran · honoured contract fingerprinted no licence file found · pointer only · 8812d1ff4a2d86e2 · report fixture: signature p:float/0/float, r:float/0/float, beta:float/0/float · returned float · level recorded 2026-09-22
comp_p metrics.py ran · honoured contract fingerprinted no licence file found · pointer only · a04366d2f95bd02b · report fixture: signature a:float/0/float, b:float/0/float · returned float · level recorded 2026-09-24
comp_r metrics.py ran · honoured contract no licence file found · pointer only · 9602afb1b61091f7 · report fixture: signature c:Tensor/0/float, g:Tensor/0/float · returned Tensor · level recorded 2026-09-24
edit_graph metrics.py ran · our draft was wrong no licence file found · pointer only · 6e7fcf15760cef34 · report fixture: signature levi_matrix:list/1/list, backpointers:dict/0/? · returned tuple · level recorded 2026-09-24
equals_ignore_whitespace_casing metrics.py ran · violated contract fingerprinted no licence file found · pointer only · 98662233319d385a · report fixture: signature a:str/0/str, b:str/0/str · returned bool · level recorded 2026-09-24
get_distance in 25 papers metrics.py ran · honoured contract no licence file found · pointer only · 10b705f17bd01668 · report fixture: signature dist:dict/0/?, v1:int/0/int, v2:int/0/int · returned float · level recorded 2026-09-24
get_ref metrics.py ran · our draft was wrong no licence file found · pointer only · 41ef3246fa487f1d · report fixture: signature edits:list/1/tuple, src:str/0/str · returned str · level recorded 2026-09-24
levenshtein_matrix metrics.py ran · our draft was wrong fingerprinted no licence file found · pointer only · b6ac128b4b5ce1c0 · report fixture: signature first:str/0/str, second:str/0/str, cost_ins:int/0/int, cost_del:int/0/int, cost_sub:int/0/int · returned tuple · level recorded 2026-09-22
load_annotation in 6 papers metrics.py ran · our draft was wrong no licence file found · pointer only · b8a21170214bed29 · report fixture: signature gold_file:str/0/str · returned tuple · level recorded 2026-09-22
matchEdit metrics.py ran · violated contract no licence file found · pointer only · 132e8a705869c3fa · report fixture: signature e:list/1/str, g:list/1/str, ignore_whitespace_casing:bool/0/bool · returned bool · level recorded 2026-09-24
matchSeq metrics.py ran · our draft was wrong no licence file found · pointer only · 83246c9a671abeec · report fixture: signature editSeq:list/1/tuple, gold_edits:list/1/tuple, ignore_whitespace_casing:bool/0/bool, verbose:bool/0/bool · returned list · level recorded 2026-09-24
merge_edits metrics.py ran · our draft was wrong no licence file found · pointer only · 74c31a434f3b4be0 · report fixture: signature e1:tuple/1/str, e2:tuple/1/str, joiner:str/0/str · returned tuple · level recorded 2026-09-24
merge_graph metrics.py ran · our draft was wrong no licence file found · pointer only · 875d2b0093f14f77 · report fixture: signature V1:list/1/int, V2:list/1/int, E1:list/1/tuple, E2:list/1/tuple, dist1:dict/0/?, dist2:dict/0/?, edits1:dict/0/?, edits2:dict/0/? · returned tuple · level recorded 2026-09-24
paragraphs metrics.py ran · our draft was wrong no licence file found · pointer only · a023f2f843d836b2 · report fixture: signature lines:list/1/str, is_separator:callable/0/?, joiner:callable/0/? · returned generator · level recorded 2026-09-24
process_edits metrics.py ran · our draft was wrong no licence file found · pointer only · e8cf8b92ebfb0e39 · report fixture: signature edits:list/1/list, dt:bool/0/bool, ds:bool/0/bool, single:bool/0/bool, filt:NoneType/0/?, multi:bool/0/bool, cse:bool/0/bool · returned dict · level recorded 2026-09-24
set_weights metrics.py ran · our draft was wrong no licence file found · pointer only · 623c7d0f98224e0f · report fixture: signature E:list/1/str, dist:dict/0/?, edits:dict/0/?, gold_edits:list/1/tuple, verbose:bool/0/bool, very_verbose:bool/0/bool · returned dict · level recorded 2026-09-24
shrinkEdit metrics.py ran · our draft was wrong no licence file found · pointer only · b122957232c466b9 · report fixture: signature edit:tuple/1/int · returned tuple · level recorded 2026-09-22
simplify_edits metrics.py ran · our draft was wrong fingerprinted no licence file found · pointer only · a45052c27ec03581 · report fixture: signature sent:str/0/str · returned list · level recorded 2026-09-22
smart_open metrics.py ran · our draft was wrong no licence file found · pointer only · 2cfd71ded15bd59d · report fixture: signature fname:str/0/str, mode:str/0/str · returned TextIOWrapper · level recorded 2026-09-24
transitive_arcs metrics.py ran · our draft was wrong no licence file found · pointer only · fac4285ee68be04f · report fixture: signature V:list/1/str, E:list/1/tuple, dist:dict/0/?, edits:dict/0/?, max_unchanged_words:int/0/int, very_verbose:bool/0/bool · returned tuple · level recorded 2026-09-24
BARTScorer metrics.py unverified no licence file found · pointer only · a131873a0a2572e5 · report
PTM2 metrics.py unverified no licence file found · pointer only · 87624a3984fe1708 · report
batch_multi_pre_rec_f1 metrics.py unverified no licence file found · pointer only · 84908af1d51cf18c · report
batch_multi_pre_rec_f1_errant metrics.py unverified no licence file found · pointer only · eb3b71d268d7c15f · report
batch_multi_pre_rec_f1_sent metrics.py unverified no licence file found · pointer only · 320d76c801f61bae · report
batch_multi_pre_rec_f1_sent_errant metrics.py unverified no licence file found · pointer only · 35662847bc7bac1f · report
compute_weight_edits metrics.py unverified no licence file found · pointer only · 0b32a42c69fe0b38 · report
errant_load_annotation metrics.py unverified no licence file found · pointer only · e3c94f36e8f91c28 · report
f1 in 57 papers metrics.py unverified no licence file found · pointer only · b8394bc429303913 · report
pre_rec_f1 metrics.py unverified no licence file found · pointer only · 2c259600beaaa31b · report

Method as code Syntology

What Syntology's graph records as this paper's claimed method, lifted into code: each assembly names its bricks (functions and classes), each with its own role and its own verification level. An assembly is not a check that the code implements the method correctly or reproduces the paper's results. Levels are each brick's own verification level in Syntology's graph (level 0: no recorded run), listed brick by brick and never combined into one level; “ran” means the brick executed on a synthesized input, not that its output is correct, and the instrument figure counts failures of Syntology's instrument, not of the code. Where a model read the code against the paper's claim, each assembly's label verdict is one model's judgement, kept in this page's JSON twin; it is not calibrated, so no card prints it.

1 assembly on this paper: 1 harvested, 0 generated.

harvested Assembly for PT-M2 (graph method: PT-M2). 31 bricks; individually on a fixture: 21 ran (of which 0 constructed an object rather than computing a result; 6 with no instrument failure: 4 honoured, 2 violated, 0 with no contract checked; 15 where Syntology's instrument failed) · 10 unverified. The assembly has not been run end-to-end.

  1. PTM2 entry · level 0 · unverified · 87624a3984fe1708
  2. BARTScorer model · level 0 · unverified · a131873a0a2572e5
  3. batch_multi_pre_rec_f1 helper · level 0 · unverified · 84908af1d51cf18c
  4. batch_multi_pre_rec_f1_errant helper · level 0 · unverified · eb3b71d268d7c15f
  5. batch_multi_pre_rec_f1_sent helper · level 0 · unverified · 320d76c801f61bae
  6. batch_multi_pre_rec_f1_sent_errant helper · level 0 · unverified · 35662847bc7bac1f
  7. errant_load_annotation helper · level 0 · unverified · e3c94f36e8f91c28
  8. f1 helper · level 0 · unverified · b8394bc429303913
  9. load_annotation helper · level 1 · ran · our draft was wrong · b8a21170214bed29
  10. smart_open helper · level 1 · ran · our draft was wrong · 2cfd71ded15bd59d
  11. best_edit_seq_bf helper · level 1 · ran · our draft was wrong · 6685898ec6c99ed8
  12. comp_f1 helper · level 1 · ran · honoured contract · 8812d1ff4a2d86e2
  13. comp_p helper · level 1 · ran · honoured contract · a04366d2f95bd02b
  14. comp_r helper · level 1 · ran · honoured contract · 9602afb1b61091f7
  15. compute_weight_edits helper · level 0 · unverified · 0b32a42c69fe0b38
  16. edit_graph helper · level 1 · ran · our draft was wrong · 6e7fcf15760cef34
  17. equals_ignore_whitespace_casing helper · level 1 · ran · violated contract · 98662233319d385a
  18. levenshtein_matrix helper · level 1 · ran · our draft was wrong · b6ac128b4b5ce1c0
  19. matchSeq helper · level 1 · ran · our draft was wrong · 83246c9a671abeec
  20. merge_graph helper · level 1 · ran · our draft was wrong · 875d2b0093f14f77
  21. set_weights helper · level 1 · ran · our draft was wrong · 623c7d0f98224e0f
  22. shrinkEdit helper · level 1 · ran · our draft was wrong · b122957232c466b9
  23. transitive_arcs helper · level 1 · ran · our draft was wrong · fac4285ee68be04f
  24. process_edits helper · level 1 · ran · our draft was wrong · e8cf8b92ebfb0e39
  25. simplify_edits helper · level 1 · ran · our draft was wrong · a45052c27ec03581
  26. pre_rec_f1 helper · level 0 · unverified · 2c259600beaaa31b
  27. paragraphs helper · level 1 · ran · our draft was wrong · a023f2f843d836b2
  28. get_ref helper · level 1 · ran · our draft was wrong · 41ef3246fa487f1d
  29. matchEdit helper · level 1 · ran · violated contract · 132e8a705869c3fa
  30. get_distance helper · level 1 · ran · honoured contract · 10b705f17bd01668
  31. merge_edits helper · level 1 · ran · our draft was wrong · 74c31a434f3b4be0

Each brick ends with its code_sha256, Syntology's identity for that exact code; click one to copy the get_code call that returns it. For agents, on Syntology's MCP service: assembly(arxiv_id="2211.01635") returns these assemblies with each brick's role and wiring (how to connect).

Tasks

Grammatical Error CorrectionMachine TranslationSentenceText Summarization

Results from the paper archive 2025-07-28

No leaderboard rows for this paper in the archive.

Report a problem or propose a change · a person checks every report against the paper or source before anything changes; decisions are listed on /corrections