Papers › Longformer: The Long-Document Transformer
Longformer: The Long-Document Transformer
Iz Beltagy, Matthew E. Peters, Arman Cohan
Transformer-based models are unable to process long sequences due to their self-attention operation, which scales quadratically with the sequence length. To address this limitation, we introduce the Longformer with an attention mechanism that scales linearly with sequence length, making it easy to process documents of thousands of tokens or longer. Longformer's attention mechanism is a drop-in replacement for the standard self-attention and combines a local windowed attention with a task motivated global attention. Following prior work on long-sequence transformers, we evaluate Longformer on character-level language modeling and achieve state-of-the-art results on text8 and enwik8. In contrast to most prior work, we also pretrain Longformer and finetune it on a variety of downstream tasks. Our pretrained Longformer consistently outperforms RoBERTa on long document tasks and sets new state-of-the-art results on WikiHop and TriviaQA. We finally introduce the Longformer-Encoder-Decoder (LED), a Longformer variant for supporting long document generative sequence-to-sequence tasks, and demonstrate its effectiveness on the arXiv summarization dataset.
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):
get_harvested_code_for_paper(arxiv_id="2004.05150")get_code(code_sha256="…")with a sample'scode_sha256from the list belowget_citation_path(paper_1="2004.05150", paper_2="…")with another paper's arXiv id or titleget_concepts_for_paper(arxiv_id="2004.05150")
Code
Syntology 35 code samples harvested from 7 repositories linked to this paper: 22 ran (of which 4 constructed an object rather than computing a result; 14 with no instrument failure: 1 honoured, 1 violated, 12 with no contract checked; 8 where Syntology's instrument failed) · 13 unverified. Where Syntology's instrument failed: 8 where our contract draft was wrong; failures of Syntology's instrument, not of the code.
By repository: official repository: 11 samples from 1 repository, 9 ran (1 with no instrument failure; 8 where Syntology's instrument failed); community (archive-listed): 24 samples from 6 repositories, 13 ran (13 with no instrument failure; 0 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.
No sample is linked to this paper from 6 of its repositories that Syntology harvested for other papers:
- han-shi/SparseBERT, community (archive-listed): harvested for 3 other papers, listed under the repository
- huggingface/transformers, community (archive-listed): harvested for 4 other papers, listed under the repository
- kit-mrt/red-motion, community (archive-listed): harvested for another paper, listed under the repository
- microsoft/dialoglm, community (archive-listed): harvested for another paper, listed under the repository
- mim-solutions/bert_for_longer_texts, community (archive-listed): harvested for another paper, listed under the repository
- mistralai/mistral-src, community (archive-listed): harvested for another paper, listed under the repository
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.
22 repositories listed; official and paper-mentioned ones first.
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.
A paper named beside a repository with no sample linked to this paper is shown with that paper's own run record, not this paper's: “ran” means executed on a synthesized input, not that the code is correct, and “where Syntology's instrument failed” counts failures of Syntology's instrument, not of the code. Papers are listed in arXiv-id order, at most three per repository.
Code Syntology ran Syntology
35 samples harvested: 22 ran (of which 4 constructed an object rather than computing a result; 14 with no instrument failure: 1 honoured, 1 violated, 12 with no contract checked; 8 where Syntology's instrument failed) · 13 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.
Licence: 5 of the 35 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 7 repositories linked to this paper, official or community; the list below is grouped by repository, each group headed by its name and what it is to the paper. “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”. For 2 samples that ran, the graph records a signature this page does not show: for a class the fixture built without calling, the harness called the object with inputs it made up itself, so the recorded signature is the harness's probe, not the fixture's arguments. “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. A census tag appears when Syntology's census ran implementations with that name on one shared input per bucket: it gives how many were compared, from how many papers, and how many distinct outputs they produced, and names this sample's bucket when this sample was one of them.
b3689e24866300ad · report
fixture: signature x:Tensor/3/float, w:int/0/int · returned Tensor · level recorded 2026-09-22 215ac6dd58d2d468 · report
fixture: signature w:int/0/int, d:int/0/int, autoregressive:bool/0/bool, device:str/0/str · returned tuple · level recorded 2026-09-22 1a6f3f01a17a0b4c · report
fixture: signature seq_len:int/0/int, w:int/0/int, d:int/0/int · returned Tensor · level recorded 2026-09-22 54c6929644249d31 · report
fixture: signature x:Tensor/4/float, direction:tuple/1/int, padding_value:float/0/float · returned Tensor · level recorded 2026-09-22 755bf82d9fe23d4b · report
fixture: signature x:Tensor/4/float, padding_value:float/0/float · returned Tensor · level recorded 2026-09-22 4261f70837b75c76 · report
fixture: signature prob:Tensor/4/float, v:Tensor/4/float, w:int/0/int · returned Tensor · level recorded 2026-09-22 1b7525edcd068b10 · report
fixture: signature q:Tensor/4/float, k:Tensor/4/float, w:int/0/int, padding_value:float/0/float · returned Tensor · level recorded 2026-09-22 1273524f034ce151 · report
fixture: signature prob:Tensor/5/float, v:Tensor/4/float, w:int/0/int · returned Tensor · level recorded 2026-09-22 3d4e46172b26669c · report
fixture: signature q:Tensor/4/float, k:Tensor/4/float, w:int/0/int, padding_value:float/0/float · returned Tensor · level recorded 2026-09-22 c1eb691dab111921 · report
bec22b3675334717 · report
62964c608dfd3d2c · report
fixture: signature tensors:list/1/Tensor · returned Tensor · level recorded 2026-09-24 9c78ceb4210479cf · report
fixture: signature lprobs:Tensor/2/float, target:Tensor/1/int, epsilon:float/0/float, pad_id:int/0/int, ignore_index:int/0/int · returned tuple · level recorded 2026-09-24 bce6838014c76d1e · report
fixture: signature tokenizer:MockTokenizer/0/?, special_token_substrings:list/1/str · returned MockTokenizer · level recorded 2026-09-24 8a1120aa9e015927 · report
e71d9c8a8cff7aa3 · report
fixture: signature hidden_states:Tensor/3/float · returned Tensor · level recorded 2026-09-22 ca468da77bd87495 · report
fixture: signature attn_size:int/0/int, window_size:int/0/int · returned Tensor · level recorded 2026-09-22 cc7b7bd19f2134ba · report
69ca0d84582f00bd · report
2a87d1deca0f2b1a · report
1d6037ddb5464048 · report
fixture: signature link:str/0/str · returned str · level recorded 2026-09-24 4f4f44d30b573fa8 · report
fixture: signature model:callable/0/?, trainable_only:bool/0/bool · returned int · level recorded 2026-09-24 e566c3f105753aa6 · report
fixture: signature model:callable/0/?, data_loader:list/1/tuple, device:device/0/?, apply_sigmoid:bool/0/bool, to_numpy:bool/0/bool · returned tuple · level recorded 2026-09-24 e71dea05795c03bf · report
fixture: signature idx:ndarray/1/int, val_size:float/0/float, test_size:float/0/float, shuffle:bool/0/bool · returned tuple · level recorded 2026-09-24 be39561477eedbab · report
census: 4 implementations from 6 papers, 4 distinct outputs across 4 buckets, one shared input per bucket src/deep_malware_detection/utils.py unverified MIT (permissive) ·
e910798b9dfc1967 · report
337700a5796189df · report
3b707d0823eee4d2 · report
2893a3bddf619d82 · report
fixture: returned FeedForward · level recorded 2026-09-22 census: 13 implementations from 9 papers, 9 distinct outputs across 5 buckets, one shared input per bucket transformer/longformer.py ran · metamorphic tier: deterministic fingerprinted MIT (permissive) ·
6c89fcd0771cd674 · report
fixture: returned PositionalEncoding · level recorded 2026-09-22 0840d53a3e87bbd0 · report
fixture: returned TokenEmbedding · level recorded 2026-09-22 4d2255c33fd872a1 · report
fixture: returned TransformerEmbedding · level recorded 2026-09-22 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. Siblings are assemblies lifted from this paper in the same batch for the same method: alternative readings kept side by side; none replaces or supersedes another.
7 assemblies on this paper: 7 harvested, 0 generated, listed in 1 groups (siblings lifted in one batch form one group).
7 sibling assemblies for Longformer (harvested), lifted in the same batch on 2026-09-09: alternative readings kept side by side; none replaces or supersedes another.
harvested Assembly for Longformer (graph method: Longformer). 6 bricks; individually on a fixture: 4 ran (of which 4 constructed an object rather than computing a result; 4 with no instrument failure: 0 honoured, 0 violated, 4 with no contract checked; 0 where Syntology's instrument failed) · 2 unverified; every one of the 4 samples that ran constructed an object rather than computing a result. The assembly has not been run end-to-end.
one of 7 siblings
- Longformer entry · level 0 · unverified ·
e23d4d61b776fd56 - LongformerLayer helper · level 0 · unverified ·
56f78e8f5e291c8d - TransformerEmbedding helper · level 1 · ran ·
4d2255c33fd872a1 - FeedForward helper · level 1 · ran ·
2893a3bddf619d82 - PositionalEncoding helper · level 1 · ran ·
6c89fcd0771cd674 - TokenEmbedding helper · level 1 · ran ·
0840d53a3e87bbd0
harvested Assembly for Longformer (graph method: Longformer). 4 bricks; individually on a fixture: 3 ran (of which 1 constructed an object rather than computing a result; 1 with no instrument failure: 0 honoured, 0 violated, 1 with no contract checked; 2 where Syntology's instrument failed) · 1 unverified. The assembly has not been run end-to-end.
one of 7 siblings
- split_tokens_into_smaller_chunks entry · level 1 · ran · our draft was wrong ·
0ce3135e26fafb14 - split_overlapping helper · level 1 · ran; the contract check could not run ·
c91a4a5114716579 - check_split_parameters_consistency helper · level 0 · unverified ·
e219be299ea45b2d - InconsistentSplittingParamsException helper · level 1 · ran ·
e4b0ec678a1f3059
harvested Assembly for Longformer (graph method: Longformer). 4 bricks; individually on a fixture: 1 ran (of which 0 constructed an object rather than computing a result; 1 with no instrument failure: 0 honoured, 1 violated, 0 with no contract checked; 0 where Syntology's instrument failed) · 3 unverified. The assembly has not been run end-to-end.
one of 7 siblings
- local_1d_pattern entry · level 1 · ran · violated contract ·
ca468da77bd87495 - local_nd_pattern helper · level 0 · unverified ·
2a87d1deca0f2b1a - local_nd_distance helper · level 0 · unverified ·
69ca0d84582f00bd - _generate_nd_grid helper · level 0 · unverified ·
cc7b7bd19f2134ba
harvested Assembly for Longformer (graph method: Longformer). 11 bricks; individually on a fixture: 9 ran (of which 0 constructed an object rather than computing a result; 1 with no instrument failure: 1 honoured, 0 violated, 0 with no contract checked; 8 where Syntology's instrument failed) · 2 unverified. The assembly has not been run end-to-end.
one of 7 siblings
- LongformerSelfAttention entry · level 0 · unverified ·
c1eb691dab111921 - mask_invalid_locations helper · level 0 · unverified ·
bec22b3675334717 - sliding_chunks_matmul_pv helper · level 1 · ran · our draft was wrong ·
4261f70837b75c76 - sliding_chunks_matmul_qk helper · level 1 · ran · our draft was wrong ·
1b7525edcd068b10 - sliding_chunks_no_overlap_matmul_pv helper · level 1 · ran · our draft was wrong ·
1273524f034ce151 - sliding_chunks_no_overlap_matmul_qk helper · level 1 · ran · our draft was wrong ·
3d4e46172b26669c - _get_invalid_locations_mask helper · level 1 · ran · our draft was wrong ·
215ac6dd58d2d468 - _skew2 helper · level 1 · ran · our draft was wrong ·
755bf82d9fe23d4b - _chunk helper · level 1 · ran · our draft was wrong ·
b3689e24866300ad - _skew helper · level 1 · ran · our draft was wrong ·
54c6929644249d31 - _get_invalid_locations_mask_fixed_dilation helper · level 1 · ran · honoured contract ·
1a6f3f01a17a0b4c
harvested Assembly for Longformer (graph method: Longformer). 11 bricks; individually on a fixture: 9 ran (of which 0 constructed an object rather than computing a result; 1 with no instrument failure: 1 honoured, 0 violated, 0 with no contract checked; 8 where Syntology's instrument failed) · 2 unverified. The assembly has not been run end-to-end.
one of 7 siblings
- LongformerSelfAttention entry · level 0 · unverified ·
3b707d0823eee4d2 - mask_invalid_locations helper · level 0 · unverified ·
bec22b3675334717 - sliding_chunks_matmul_pv helper · level 1 · ran · our draft was wrong ·
4261f70837b75c76 - sliding_chunks_matmul_qk helper · level 1 · ran · our draft was wrong ·
1b7525edcd068b10 - sliding_chunks_no_overlap_matmul_pv helper · level 1 · ran · our draft was wrong ·
1273524f034ce151 - sliding_chunks_no_overlap_matmul_qk helper · level 1 · ran · our draft was wrong ·
3d4e46172b26669c - _get_invalid_locations_mask helper · level 1 · ran · our draft was wrong ·
215ac6dd58d2d468 - _skew2 helper · level 1 · ran · our draft was wrong ·
755bf82d9fe23d4b - _chunk helper · level 1 · ran · our draft was wrong ·
b3689e24866300ad - _skew helper · level 1 · ran · our draft was wrong ·
54c6929644249d31 - _get_invalid_locations_mask_fixed_dilation helper · level 1 · ran · honoured contract ·
1a6f3f01a17a0b4c
harvested Assembly for Longformer (graph method: Longformer). 2 bricks; individually on a fixture: 1 ran (of which 0 constructed an object rather than computing a result; 1 with no instrument failure: 0 honoured, 0 violated, 1 with no contract checked; 0 where Syntology's instrument failed) · 1 unverified. The assembly has not been run end-to-end.
one of 7 siblings
- LongformerClass entry · level 0 · unverified ·
083425b2586f0566 - LongformerClassificationHead model · level 1 · ran ·
e71d9c8a8cff7aa3
harvested Assembly for Longformer (graph method: Longformer). 1 brick; individually on a fixture: 0 ran · 1 unverified. The assembly has not been run end-to-end.
one of 7 siblings
- BertSelfAttention entry · level 0 · unverified ·
0cddc969b8e047cc
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="2004.05150") returns these assemblies with each brick's role and wiring (how to connect).
Tasks
Results from the paper archive 2025-07-28
| Task | Dataset | Model | Metric | Value | Rank at snapshot | Leaderboard | Report |
|---|---|---|---|---|---|---|---|
| Language Modelling | Hutter Prize | Longformer Large | Bit per Character (BPC) | 0.99 | #5 of 18 | Archive leaderboard | report |
| Language Modelling | Hutter Prize | Longformer Large | Number of params | 102M | #5 of 18 | Archive leaderboard | report |
| Language Modelling | Hutter Prize | Longformer Small | Bit per Character (BPC) | 1.00 | #6 of 18 | Archive leaderboard | report |
| Language Modelling | Hutter Prize | Longformer Small | Number of params | 41M | #6 of 18 | Archive leaderboard | report |
| Language Modelling | enwik8 | Longformer (30 layers, h=512) | Bit per Character (BPC) | 0.99 | #13 of 42 | Archive leaderboard | report |
| Language Modelling | enwik8 | Longformer (30 layers, h=512) | Number of params | 102M | #13 of 42 | Archive leaderboard | report |
| Language Modelling | enwik8 | Longformer (12 layers, h=512) | Bit per Character (BPC) | 1.00 | #18 of 42 | Archive leaderboard | report |
| Language Modelling | enwik8 | Longformer (12 layers, h=512) | Number of params | 41M | #18 of 42 | Archive leaderboard | report |
| Question Answering | WikiHop | Longformer-large | Test | 81.9 | #2 of 9 | Archive leaderboard | report |
Ranks are positions in the archive's leaderboards as they stood at the 2025-07-28 snapshot. Results published since then are not among these rows, so a rank here is not a current standing.
Methods
Introduced by this paper: Longformer
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