Papers › Block-Recurrent Transformers

Block-Recurrent Transformers

11 Mar 2022arXiv:2203.07852archive 2025-07-28

DeLesley Hutchins, Imanol Schlag, Yuhuai Wu, Ethan Dyer, Behnam Neyshabur

We introduce the Block-Recurrent Transformer, which applies a transformer layer in a recurrent fashion along a sequence, and has linear complexity with respect to sequence length. Our recurrent cell operates on blocks of tokens rather than single tokens during training, and leverages parallel computation within a block in order to make efficient use of accelerator hardware. The cell itself is strikingly simple. It is merely a transformer layer: it uses self-attention and cross-attention to efficiently compute a recurrent function over a large set of state vectors and tokens. Our design was inspired in part by LSTM cells, and it uses LSTM-style gates, but it scales the typical LSTM cell up by several orders of magnitude. Our implementation of recurrence has the same cost in both computation time and parameter count as a conventional transformer layer, but offers dramatically improved perplexity in language modeling tasks over very long sequences. Our model out-performs a long-range Transformer XL baseline by a wide margin, while running twice as fast. We demonstrate its effectiveness on PG19 (books), arXiv papers, and GitHub source code. Our code has been released as open source.

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

No sample from the archive's official repository (google-research/meliad) is linked to this paper; the 19 samples below come from 2 community repositories, and 4 more are identical to code Syntology harvested from another paper's repository. Syntology harvested code from google-research/meliad for another paper.

google-research/meliad: Syntology's graph links these samples to Memorizing Transformers (arXiv:2203.08913; that paper's own run record: 23 ran (of which 6 constructed an object rather than computing a result; 18 with no instrument failure: 3 honoured, 1 violated, 14 with no contract checked; 5 where Syntology's instrument failed) · 11 unverified); nothing checks that they implement this paper's method.

Syntology 23 code samples: 19 from 2 repositories linked to this paper, 4 identical to code harvested from another paper's repository. Across all 23: 18 ran (of which 2 constructed an object rather than computing a result; 9 with no instrument failure: 2 honoured, 3 violated, 4 with no contract checked; 9 where Syntology's instrument failed) · 5 unverified. Where Syntology's instrument failed: 8 where our contract draft was wrong; 1 where the contract check could not run; failures of Syntology's instrument, not of the code.

By repository: community (archive-listed): 19 samples from 2 repositories, 14 ran (6 with no instrument failure; 8 where Syntology's instrument failed); 4 identical to code first harvested elsewhere. 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 1 of its repositories that Syntology harvested for other papers:

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. 7 of the 18 samples that ran have no recorded date.

google-research/meliad officialmentioned in papermentioned on GitHubjaxApache-2.0Syntology: no sample linked to this paper (harvested for another paper). Syntology's graph links these samples to Memorizing Transformers (arXiv:2203.08913; that paper's own run record: 23 ran (of which 6 constructed an object rather than computing a result; 18 with no instrument failure: 3 honoured, 1 violated, 14 with no contract checked; 5 where Syntology's instrument failed) · 11 unverified); nothing checks that they implement this paper's method. report
dashstander/block-recurrent-transformer mentioned on GitHubpytorchSyntology: 3 samples, 3 ran (2 instrument-failed) · 0 unverified report
lucidrains/block-recurrent-transformer-pytorch mentioned on GitHubpytorchSyntology: 16 samples, 11 ran (6 instrument-failed) · 5 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.

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

23 samples harvested: 18 ran (of which 2 constructed an object rather than computing a result; 9 with no instrument failure: 2 honoured, 3 violated, 4 with no contract checked; 9 where Syntology's instrument failed) · 5 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.

2ran · honoured contract
3ran · violated contract
8ran · our draft was wrong
1ran; the contract check could not run
4ran
5unverified

Licence: 4 of the 23 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 2 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. 4 samples are listed last, in a group without a repository. “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 1 sample 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.

dashstander/block-recurrent-transformer community (archive-listed) 3 samples, 3 ran (2 instrument-failed) · 0 unverified
apply_rotary_pos_emb in 393 papers
census: 14 implementations from 15 papers, 8 distinct outputs across 7 buckets, one shared input per bucket this sample: bucket 2 of 7, arg 1: rank 3, kind float, dtype float32 · arg 2: rank 3, kind float, dtype float32
block_recurrent_transformer/transformer.py ran · our draft was wrong MIT (permissive) · a0a60756432aafb6 · report fixture: signature t:Tensor/3/float, freqs:Tensor/3/float · returned Tensor · level recorded, date not recorded
cast_tuple in 58 papers block_recurrent_transformer/transformer.py ran · violated contract fingerprinted MIT (permissive) · 75e2de7f7799802d · report fixture: signature val:int/0/int, num:int/0/int · returned tuple · level recorded, date not recorded
rotate_half in 424 papers
census: 24 implementations from 96 papers, 4 distinct outputs across 2 buckets, one shared input per bucket this sample: bucket 2 of 2, arg 1: rank 3, kind float, dtype float32
block_recurrent_transformer/transformer.py ran · our draft was wrong fingerprinted MIT (permissive) · 6b01a575e36ad18c · report fixture: signature x:Tensor/3/float · returned Tensor · level recorded, date not recorded
lucidrains/block-recurrent-transformer-pytorch community (archive-listed) 16 samples, 11 ran (6 instrument-failed) · 5 unverified
is_empty in 15 papers block_recurrent_transformer_pytorch/block_recurrent_transformer_pytorch.py ran · our draft was wrong MIT (permissive) · 039ac61b2d11642b · report fixture: signature t:Tensor/1/int · returned bool · level recorded, date not recorded
FeedForward in 198 papers block_recurrent_transformer_pytorch/block_recurrent_transformer_pytorch.py ran · our draft was wrong MIT (permissive) · b0bf82c667f91301 · report fixture: signature dim:int/0/int, mult:int/0/int · returned Sequential · level recorded 2026-09-22
GEGLU in 20 papers
census: 12 implementations from 12 papers, 1 distinct output on one shared input this sample: bucket 1 of 1, arg 1: rank 2, kind float, dtype float32
block_recurrent_transformer_pytorch/block_recurrent_transformer_pytorch.py ran · metamorphic tier: invariant fingerprinted MIT (permissive) · c5c74119405bfc64 · report fixture: signature x:Tensor/2/float · returned Tensor · level recorded 2026-09-22
MemoryManager block_recurrent_transformer_pytorch/block_recurrent_transformer_pytorch.py ran MIT (permissive) · 778a8ddac6fbf372 · report fixture: signature past_memories:list/1/Tensor, new_memories:list/1/Tensor · returned list · level recorded 2026-09-24
RotaryEmbedding in 27 papers block_recurrent_transformer_pytorch/block_recurrent_transformer_pytorch.py ran MIT (permissive) · c779f10c971c5655 · report fixture: returned RotaryEmbedding · level recorded 2026-09-22
all_unique block_recurrent_transformer_pytorch/block_recurrent_transformer_pytorch.py ran · our draft was wrong fingerprinted MIT (permissive) · 2b23ad0c970e5ae8 · report fixture: signature arr:list/1/int · returned bool · level recorded 2026-09-22
apply_rotary_pos_emb in 393 papers
census: 14 implementations from 15 papers, 8 distinct outputs across 7 buckets, one shared input per bucket this sample: bucket 1 of 7, arg 1: rank 3, kind float, dtype float32 · arg 2: rank 2, kind float, dtype float32
block_recurrent_transformer_pytorch/block_recurrent_transformer_pytorch.py ran · our draft was wrong fingerprinted MIT (permissive) · a1f5fb7641bf7ce3 · report fixture: signature t:Tensor/3/float, pos:Tensor/2/float, scale:float/0/float · returned Tensor · level recorded 2026-09-22
cast_tuple in 58 papers block_recurrent_transformer_pytorch/block_recurrent_transformer_pytorch.py ran · honoured contract fingerprinted MIT (permissive) · c1b68cdd9e73316a · report fixture: signature t:int/0/int, length:int/0/int · returned tuple · level recorded 2026-09-22
gumbel_sample in 15 papers
census: 6 implementations from 5 papers, 2 distinct outputs on one shared input this sample: bucket 1 of 1, arg 1: rank 2, kind float, dtype float32
block_recurrent_transformer_pytorch/block_recurrent_transformer_pytorch.py ran; the contract check could not run (RAISES) MIT (permissive) · 41e347331d03a86c · report fixture: signature t:Tensor/2/float, temperature:float/0/float, dim:int/0/int · returned Tensor · level recorded 2026-09-22
top_k in 17 papers
census: 7 implementations from 9 papers, 5 distinct outputs on one shared input this sample: bucket 1 of 1, arg 1: rank 2, kind float, dtype float32
block_recurrent_transformer_pytorch/block_recurrent_transformer_pytorch.py ran · our draft was wrong fingerprinted MIT (permissive) · 39f3cec3c6be459f · report fixture: signature logits:Tensor/2/float, thres:float/0/float · returned Tensor · level recorded 2026-09-22
AttentionBlock in 69 papers block_recurrent_transformer_pytorch/block_recurrent_transformer_pytorch.py unverified MIT (permissive) · 46353ab6ce1236e2 · report
BlockRecurrentTransformer block_recurrent_transformer_pytorch/block_recurrent_transformer_pytorch.py unverified MIT (permissive) · 8c4ac9d79f99be7d · report
StateContainer block_recurrent_transformer_pytorch/block_recurrent_transformer_pytorch.py unverified MIT (permissive) · 9278e73f426f6715 · report
safe_cat block_recurrent_transformer_pytorch/block_recurrent_transformer_pytorch.py unverified MIT (permissive) · 54c65db65b0fbb51 · report
Repository not located 4 samples, 4 ran (1 instrument-failed) · 0 unverified
4 samples are identical to code Syntology harvested from another paper's repository; this paper's own copy was not located, so no repository is named and no licence is recorded.
compact in 8 papers repository not located ran · honoured contract fingerprinted licence of this copy not recorded · 4a06bf1ac86f840a · report fixture: signature arr:list/1/int · returned list · level recorded 2026-09-22
default in 430 papers repository not located ran · violated contract fingerprinted licence of this copy not recorded · 60fff7c3c400d7ff · report fixture: signature val:int/0/int, d:int/0/int · returned int · level recorded, date not recorded
eval_decorator in 15 papers repository not located ran · our draft was wrong licence of this copy not recorded · 5444a7d75d38878d · report fixture: signature fn:callable/0/? · returned function · level recorded, date not recorded
exists in 528 papers
census: 5 implementations from 103 papers, 1 distinct output on one shared input this sample: bucket 1 of 1, arg 1: rank 1, kind int, dtype int64
repository not located ran · violated contract licence of this copy not recorded · aa5486a3650902d8 · report fixture: signature val:Tensor/1/int · returned bool · level recorded, date not recorded

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 Block-Recurrent Transformer (graph method: Block-Recurrent Transformer). 30 bricks; individually on a fixture: 23 ran (of which 2 constructed an object rather than computing a result; 12 with no instrument failure: 4 honoured, 4 violated, 4 with no contract checked; 11 where Syntology's instrument failed) · 7 unverified. The assembly has not been run end-to-end.

  1. BlockRecurrentTransformer entry · level 0 · unverified · 8c4ac9d79f99be7d
  2. AttentionBlock model · level 0 · unverified · 46353ab6ce1236e2
  3. FeedForward model · level 1 · ran · our draft was wrong · b0bf82c667f91301
  4. LayerNorm model · level 1 · ran · f6fa422c7a33d41e
  5. MemoryManager model · level 1 · ran · 778a8ddac6fbf372
  6. RotaryEmbedding model · level 1 · ran · c779f10c971c5655
  7. all_unique helper · level 1 · ran · our draft was wrong · 2b23ad0c970e5ae8
  8. default helper · level 1 · ran · violated contract · 60fff7c3c400d7ff
  9. divisible_by helper · level 2 · ran · violated contract · 856d86b39c6005b4
  10. eval_decorator helper · level 1 · ran · our draft was wrong · 5444a7d75d38878d
  11. exists helper · level 1 · ran · violated contract · aa5486a3650902d8
  12. gumbel_sample helper · level 1 · ran; the contract check could not run (RAISES) · 41e347331d03a86c
  13. top_k helper · level 1 · ran · our draft was wrong · 39f3cec3c6be459f
  14. Attention helper · level 0 · unverified · 20f6b4952e3785e4
  15. StateContainer model · level 0 · unverified · 9278e73f426f6715
  16. GEGLU helper · level 1 · ran · c5c74119405bfc64
  17. cast_tuple helper · level 1 · ran · honoured contract · c1b68cdd9e73316a
  18. is_empty helper · level 2 · ran · our draft was wrong · 039ac61b2d11642b
  19. safe_cat helper · level 0 · unverified · 54c65db65b0fbb51
  20. gumbel_noise helper · level 1 · ran · honoured contract · d85a55ec5c2984ef
  21. Attend model · level 0 · unverified · a91304b317adcfe6
  22. apply_rotary_pos_emb helper · level 1 · ran · our draft was wrong · a1f5fb7641bf7ce3
  23. l2norm helper · level 1 · ran · our draft was wrong · 3eb58dae620046d3
  24. compact helper · level 1 · ran · honoured contract · 4a06bf1ac86f840a
  25. log helper · level 2 · ran · our draft was wrong · b5472673332de78e
  26. and_reduce helper · level 1 · ran · violated contract · 5e1eae834b42ab87
  27. pack_one helper · level 0 · unverified · 86ca4a3f9ab20a35
  28. unpack_one helper · level 1 · ran · our draft was wrong · a0902e24bf5299e7
  29. rotate_half helper · level 1 · ran · honoured contract · 437c7011506995a3
  30. once helper · level 1 · ran · our draft was wrong · 2ae574072ec6f776

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="2203.07852") returns these assemblies with each brick's role and wiring (how to connect).

Tasks

Language ModelingLanguage Modelling

Results from the paper archive 2025-07-28

No leaderboard rows for this paper in the archive.

Methods

Absolute Position EncodingsAdamAttentionBPEDense ConnectionsDropoutLSTMLabel SmoothingLayer NormalizationLinear LayerMulti-Head AttentionPosition-Wise Feed-Forward LayerResidual ConnectionSigmoid ActivationSoftmaxTanh ActivationTransformer

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