Papers › Improving Massively Multilingual Neural Machine Translation and Zero-Shot Translation

Improving Massively Multilingual Neural Machine Translation and Zero-Shot Translation

24 Apr 2020ACL 2020 6arXiv:2004.11867archive 2025-07-28

Biao Zhang, Philip Williams, Ivan Titov, Rico Sennrich

Massively multilingual models for neural machine translation (NMT) are theoretically attractive, but often underperform bilingual models and deliver poor zero-shot translations. In this paper, we explore ways to improve them. We argue that multilingual NMT requires stronger modeling capacity to support language pairs with varying typological characteristics, and overcome this bottleneck via language-specific components and deepening NMT architectures. We identify the off-target translation issue (i.e. translating into a wrong target language) as the major source of the inferior zero-shot performance, and propose random online backtranslation to enforce the translation of unseen training language pairs. Experiments on OPUS-100 (a novel multilingual dataset with 100 languages) show that our approach substantially narrows the performance gap with bilingual models in both one-to-many and many-to-many settings, and improves zero-shot performance by ~10 BLEU, approaching conventional pivot-based methods.

PaperPDFConference PDFCodeCode 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 13 code samples harvested from 1 repository linked to this paper: 2 ran (of which 0 constructed an object rather than computing a result; 0 with no instrument failure: 0 honoured, 0 violated, 0 with no contract checked; 2 where Syntology's instrument failed) · 11 unverified. Where Syntology's instrument failed: 2 where the contract check could not run; failures of Syntology's instrument, not of the code. Every run on this page is a failure of Syntology's instrument, not of the code.

By repository: official repository: 13 samples from 1 repository, 2 ran (0 with no instrument failure; 2 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 on 2026-09-22 (UTC): the dates Syntology's graph wrote each sample's verification level, not when the code executed.

EdinburghNLP/opus-100-corpus officialmentioned in paperSyntology: not harvested report
bzhangGo/zero officialmentioned in papertfSyntology: 13 samples, 2 ran (2 instrument-failed) · 11 unverified report
YuweiYin/HLT-MT mentioned on GitHubpytorchSyntology: not harvested 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

13 samples harvested: 2 ran (of which 0 constructed an object rather than computing a result; 0 with no instrument failure: 0 honoured, 0 violated, 0 with no contract checked; 2 where Syntology's instrument failed) · 11 unverified. “Where Syntology's instrument failed” counts failures of Syntology's instrument, not of the code. Every run on this page is a failure 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.

2ran; the contract check could not run
11unverified

Licence: 0 of the 13 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 bzhangGo/zero. “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.

bzhangGo/zero official repository 13 samples, 2 ran (2 instrument-failed) · 11 unverified
combine_heads in 6 papers models/transformer_l0drop.py ran; the contract check could not run (RAISES) · metamorphic tier: invariant fingerprinted BSD-3-Clause (permissive) · 9faf4852d7f74a91 · report fixture: signature inputs:EagerTensor/4/float, name:str/0/str · returned EagerTensor · level recorded 2026-09-22
split_heads in 10 papers models/transformer_l0drop.py ran; the contract check could not run (RAISES) · metamorphic tier: invariant fingerprinted BSD-3-Clause (permissive) · 688a6272dc140cd0 · report fixture: signature inputs:EagerTensor/3/float, num_heads:int/0/int, name:NoneType/0/? · returned EagerTensor · level recorded 2026-09-22
add_timing_signal in 5 papers models/transformer_l0drop.py unverified BSD-3-Clause (permissive) · 4b52e600cf72c348 · report
attention_bias models/transformer_l0drop.py unverified BSD-3-Clause (permissive) · b88b2db9eb46e73d · report
decoder in 19 papers models/transformer_l0drop.py unverified BSD-3-Clause (permissive) · c01bf5a6fe99ddfb · report
dot_attention in 7 papers models/transformer_l0drop.py unverified BSD-3-Clause (permissive) · b04d0b5aeab65a69 · report
encoder in 30 papers models/transformer_l0drop.py unverified BSD-3-Clause (permissive) · dd6b86752ec524cd · report
extract_encodes models/transformer_l0drop.py unverified BSD-3-Clause (permissive) · 5b2126720aede56f · report
ffn_layer models/transformer_l0drop.py unverified BSD-3-Clause (permissive) · 36e7645b57365d0d · report
layer_norm in 32 papers models/transformer_l0drop.py unverified BSD-3-Clause (permissive) · 9dcd5100e4896669 · report
linear in 88 papers models/transformer_l0drop.py unverified BSD-3-Clause (permissive) · c9710c0dc3de38fe · report
residual_fn models/transformer_l0drop.py unverified BSD-3-Clause (permissive) · a58e7c7393bd5449 · report
train_fn models/transformer_l0drop.py unverified BSD-3-Clause (permissive) · 81602c4d87412b16 · 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 random online backtranslation (graph method: random online backtranslation). 13 bricks; individually on a fixture: 2 ran (of which 0 constructed an object rather than computing a result; 0 with no instrument failure: 0 honoured, 0 violated, 0 with no contract checked; 2 where Syntology's instrument failed) · 11 unverified. The assembly has not been run end-to-end.

  1. train_fn entry · level 0 · unverified · 81602c4d87412b16
  2. decoder model · level 0 · unverified · c01bf5a6fe99ddfb
  3. encoder model · level 0 · unverified · dd6b86752ec524cd
  4. add_timing_signal helper · level 0 · unverified · 4b52e600cf72c348
  5. attention_bias helper · level 0 · unverified · b88b2db9eb46e73d
  6. dot_attention model · level 0 · unverified · b04d0b5aeab65a69
  7. extract_encodes model · level 0 · unverified · 5b2126720aede56f
  8. ffn_layer helper · level 0 · unverified · 36e7645b57365d0d
  9. layer_norm helper · level 0 · unverified · 9dcd5100e4896669
  10. linear helper · level 0 · unverified · c9710c0dc3de38fe
  11. residual_fn helper · level 0 · unverified · a58e7c7393bd5449
  12. combine_heads helper · level 1 · ran; the contract check could not run (RAISES) · 9faf4852d7f74a91
  13. split_heads helper · level 1 · ran; the contract check could not run (RAISES) · 688a6272dc140cd0

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="2004.11867") returns these assemblies with each brick's role and wiring (how to connect).

Tasks

Machine TranslationNMTTranslation

Datasets

Introduced by this paper, per the archive.

OPUS-100

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