Papers › Gated Delta Networks: Improving Mamba2 with Delta Rule

Gated Delta Networks: Improving Mamba2 with Delta Rule

9 Dec 2024arXiv:2412.06464archive 2025-07-28

Songlin Yang, Jan Kautz, Ali Hatamizadeh

Linear Transformers have gained attention as efficient alternatives to standard Transformers, but their performance in retrieval and long-context tasks has been limited. To address these limitations, recent work has explored two distinct mechanisms: gating for adaptive memory control and the delta update rule for precise memory modifications. We observe that these mechanisms are complementary: gating enables rapid memory erasure while the delta rule facilitates targeted updates. Building on this insight, we introduce the gated delta rule and develop a parallel training algorithm optimized for modern hardware. Our proposed architecture, Gated DeltaNet, consistently surpasses existing models like Mamba2 and DeltaNet across multiple benchmarks, including language modeling, common-sense reasoning, in-context retrieval, length extrapolation, and long-context understanding. We further enhance performance by developing hybrid architectures that combine Gated DeltaNet layers with sliding window attention or Mamba2 layers, achieving both improved training efficiency and superior task performance.

PaperPDFCodeCode Syntology ran

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

For agents, Syntology's MCP tool lists every function and class Syntology harvested from this paper and whether it ran (how to connect): get_harvested_code_for_paper(arxiv_id="2412.06464")

Code

Syntology Ran 5 of 13 code samples harvested from 2 repositories linked to this paper; 8 have no recorded run. Of those that ran: 1 ran · honoured contract; 4 ran with no contract checked.

By repository: official repository: 7 samples from 1 repository, 5 ran; community (archive-listed): 6 samples from 1 repository, 0 ran. The run record, sample by sample. “Ran” means executed on a synthesized input, not that the code is correct or reproduces the paper.

NVlabs/GatedDeltaNet officialmentioned on GitHubpytorchNOASSERTION report
fla-org/flash-linear-attention mentioned on GitHubpytorch report
sustcsonglin/flash-linear-attention mentioned on GitHubpytorchMIT report
torchrwkv/flash-linear-attention mentioned on GitHubpytorchMIT 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; 5 ran; 1 honoured the contract we drafted; 8 have no recorded run. Read from Syntology's graph 2026-09-24; that is when this build read the record, not when the samples ran.

1ran · honoured contract
4ran
8unverified

Licence: 7 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 2 repositories linked to this paper, official or community; each sample names its own and says which. “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. “Honoured” and “violated” refer to a contract Syntology drafted from the code itself; “our draft was wrong” and “fixture could not drive it” are failures of Syntology's instrument, not of the code.

Each sample ends with its code_sha256, Syntology's identity for that exact code. An agent fetches the stored sample with Syntology's MCP tool get_code(code_sha256="…") (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 from a repository marked official are listed first. 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 four 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; or the licence is outside the permissive list Syntology serves text under (MIT, Apache-2.0, BSD and similar). 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.

apply_rotary_emb_torch NVlabs/GatedDeltaNet/lit_gpt/rotary.py official repository ran no licence file found · pointer only · 5ec949977c334f87 · report
build_rope_cache NVlabs/GatedDeltaNet/lit_gpt/model.py official repository ran · honoured contract no licence file found · pointer only · a30c7463cb6330d7 · report
code NVlabs/GatedDeltaNet/lit_gpt/packed_dataset.py official repository ran no licence file found · pointer only · fa298642043e1282 · report
maybe_align NVlabs/GatedDeltaNet/lit_gpt/rmsnorm.py official repository ran fingerprinted no licence file found · pointer only · c6283b7a1a1785ba · report
rotate_half NVlabs/GatedDeltaNet/lit_gpt/rotary.py official repository ran fingerprinted no licence file found · pointer only · 163292f43a3c861b · report
create_block NVlabs/GatedDeltaNet/lit_gpt/model.py official repository unverified licence not identified · pointer only · cf8f6cb38ea274f7 · report
get_flops_available NVlabs/GatedDeltaNet/lit_gpt/speed_monitor.py official repository unverified no licence file found · pointer only · 3f3d25ea807f6d5c · report
ceil_div torchrwkv/flash-linear-attention/fla/ops/log_linear_attn/chunk.py community (archive-listed) unverified MIT (permissive) · 881e6b15cd91e16c · report
construct_H_matrix torchrwkv/flash-linear-attention/fla/ops/log_linear_attn/naive.py community (archive-listed) unverified MIT (permissive) · 35a90dba5e6339cb · report
construct_binary_level_mask torchrwkv/flash-linear-attention/fla/ops/log_linear_attn/chunk.py community (archive-listed) unverified MIT (permissive) · f425d46b52d30143 · report
construct_level_mask torchrwkv/flash-linear-attention/fla/ops/log_linear_attn/naive.py community (archive-listed) unverified MIT (permissive) · 2eeea785ff36049a · report
level_lut torchrwkv/flash-linear-attention/fla/ops/log_linear_attn/chunk.py community (archive-listed) unverified MIT (permissive) · 2196376162089b22 · report
segsum torchrwkv/flash-linear-attention/fla/ops/log_linear_attn/naive.py community (archive-listed) unverified MIT (permissive) · f0e873ec510a64bc · report

Tasks

Common Sense ReasoningLanguage ModelingLanguage ModellingLong-Context UnderstandingRetrieval

Results from the paper archive 2025-07-28

No leaderboard rows for this paper in the archive.

Methods

AttentionSoftmax

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