Papers › Sparse and Continuous Attention Mechanisms

Sparse and Continuous Attention Mechanisms

12 Jun 2020NeurIPS 2020 12arXiv:2006.07214archive 2025-07-28

André F. T. Martins, António Farinhas, Marcos Treviso, Vlad Niculae, Pedro M. Q. Aguiar, Mário A. T. Figueiredo

Exponential families are widely used in machine learning; they include many distributions in continuous and discrete domains (e.g., Gaussian, Dirichlet, Poisson, and categorical distributions via the softmax transformation). Distributions in each of these families have fixed support. In contrast, for finite domains, there has been recent work on sparse alternatives to softmax (e.g. sparsemax and alpha-entmax), which have varying support, being able to assign zero probability to irrelevant categories. This paper expands that work in two directions: first, we extend alpha-entmax to continuous domains, revealing a link with Tsallis statistics and deformed exponential families. Second, we introduce continuous-domain attention mechanisms, deriving efficient gradient backpropagation algorithms for alpha in {1,2}. Experiments on attention-based text classification, machine translation, and visual question answering illustrate the use of continuous attention in 1D and 2D, showing that it allows attending to time intervals and compact regions.

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.

Syntology holds a reference implementation of α-entmax extension to continuous domains, generated from the paper text (not the authors' code), at level 2 on Syntology's ladder: V2 “two implementations GENERATED from the paper's signature and spec agree numerically (1e-6) on fixture inputs” “neither of the two ever read the authors' own code, so V2 measures how unambiguously the paper specifies the method, not fidelity to the authors' implementation (measured 2026-09-11, commit 75eaa77: cross_checked_by is a second LLM on 2,628 of 2,831 served samples)” (quoted from Syntology's server card). No rung compares this code against the authors' implementation. For agents: get_reference_implementation(name='α-entmax extension to continuous domains') (how to connect); the free check have('2006.07214') needs no token.

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

Code

Syntology 15 code samples harvested from 2 repositories linked to this paper: 12 ran (of which 9 constructed an object rather than computing a result; 9 with no instrument failure: 0 honoured, 0 violated, 9 with no contract checked; 3 where Syntology's instrument failed) · 3 unverified. Where Syntology's instrument failed: 3 where our contract draft was wrong; failures of Syntology's instrument, not of the code.

By repository: official repository: 2 samples from 1 repository, 0 ran (0 with no instrument failure; 0 where Syntology's instrument failed); community (archive-listed): 13 samples from 1 repository, 12 ran (9 with no instrument failure; 3 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.

deep-spin/mcan-vqa-continuous-attention officialmentioned in papermentioned on GitHubpytorchSyntology: 2 samples, 0 ran · 2 unverified report
deep-spin/quati mentioned on GitHubpytorchSyntology: 13 samples, 12 ran (3 instrument-failed) · 1 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

15 samples harvested: 12 ran (of which 9 constructed an object rather than computing a result; 9 with no instrument failure: 0 honoured, 0 violated, 9 with no contract checked; 3 where Syntology's instrument failed) · 3 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.

3ran · our draft was wrong
9ran
3unverified

Licence: 2 of the 15 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. “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.

deep-spin/mcan-vqa-continuous-attention official repository 2 samples, 0 ran · 2 unverified
ContinuousSparsemax core/model/continuous_sparsemax.py unverified no licence file found · pointer only · 4d564215130f97e9 · report
ContinuousSparsemaxFunction core/model/continuous_sparsemax.py unverified no licence file found · pointer only · dc6bd3d88b59cbda · report
deep-spin/quati community (archive-listed) 13 samples, 12 ran (3 instrument-failed) · 1 unverified
BasisFunctions quati/modules/continuous_attention.py ran MIT (permissive) · 544ace0143134f21 · report fixture: returned BasisFunctions · level recorded 2026-09-22
ContinuousSoftmax quati/modules/continuous_attention.py ran · metamorphic tier: deterministic fingerprinted MIT (permissive) · 88d5d46a3e9f268f · report fixture: returned ContinuousSoftmax · level recorded 2026-09-22
ContinuousSoftmaxFunction quati/modules/continuous_attention.py ran MIT (permissive) · 2abd421a2409332c · report fixture: returned ContinuousSoftmaxFunction · level recorded 2026-09-22
ContinuousSparsemax quati/modules/continuous_attention.py ran · metamorphic tier: deterministic MIT (permissive) · 91ec632f03a8f06b · report fixture: returned ContinuousSparsemax · level recorded 2026-09-22
ContinuousSparsemaxFunction quati/modules/continuous_attention.py ran MIT (permissive) · 45b9c83a5ff5f057 · report fixture: returned ContinuousSparsemaxFunction · level recorded 2026-09-22
CosineBasisFunctions quati/modules/continuous_attention.py ran · metamorphic tier: deterministic MIT (permissive) · 0513653665d957f6 · report fixture: returned CosineBasisFunctions · level recorded 2026-09-22
GaussianBasisFunctions quati/modules/continuous_attention.py ran · metamorphic tier: deterministic MIT (permissive) · d4a4993b0f77f70a · report fixture: returned GaussianBasisFunctions · level recorded 2026-09-22
PowerBasisFunctions quati/modules/continuous_attention.py ran MIT (permissive) · af4a48bd715bb239 · report fixture: returned PowerBasisFunctions · level recorded 2026-09-22
SineBasisFunctions quati/modules/continuous_attention.py ran · metamorphic tier: deterministic MIT (permissive) · a041cde2266ca5ff · report fixture: returned SineBasisFunctions · level recorded 2026-09-22
add_gaussian_basis_functions quati/modules/continuous_attention.py ran · our draft was wrong MIT (permissive) · b2b0552bda1b4c2d · report fixture: signature nb_basis:int/0/int, sigmas:list/1/float, device:NoneType/0/? · returned GaussianBasisFunctions · level recorded 2026-09-22
add_power_basis_functions quati/modules/continuous_attention.py ran · our draft was wrong MIT (permissive) · 4500f4cfd4afb4ee · report fixture: signature min_d:int/0/int, max_d:int/0/int, device:NoneType/0/? · returned PowerBasisFunctions · level recorded 2026-09-22
add_wave_basis_functions quati/modules/continuous_attention.py ran · our draft was wrong fingerprinted MIT (permissive) · 7811322c90f19d30 · report fixture: signature nb_waves:int/0/int, wave_b:float/0/float, max_seq_len:int/0/int, device:NoneType/0/? · returned tuple · level recorded 2026-09-22
ContinuousAttention quati/modules/continuous_attention.py unverified MIT (permissive) · 7bc1d8632d6b41c5 · 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. For a generated assembly, an independent peer is a second implementation generated separately from the same paper text; neither implementation read the authors' code. At level 2 the two agreed numerically on the fixture's inputs; level 3 is recorded when property tests derived from the paper also passed on both, and each level 3 card prints how many its report lists. 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. A generated assembly's label was set when its code was generated for the method, and no model judged it; its card prints it as "by construction" or "not judged by a model". Where an assembly's phrase has no constructed figure, a brick that ran is not among the harvested samples this build reads, and whether a brick's run only constructed an object is read from that brick's harvested sample, so the figure was not counted; it does not say there were none.

3 assemblies on this paper: 2 harvested, 1 generated.

generated Assembly for α-entmax extension to continuous domains (graph method: α-entmax extension to continuous domains), generated from the paper text, not the repository's code. 1 brick; individually on a fixture: 1 ran (1 with no instrument failure: 0 honoured, 0 violated, 1 with no contract checked; 0 where Syntology's instrument failed) · 0 unverified. Ran end-to-end against an independent peer (level 2).

label matches by construction: the code was generated for this method (not judged by a model)

  1. continuous_softmax_attention_backward entry · level 2 · ran · de96465eb76a552fc41e88b6f1fa1e03b27f5ff269e2830b8e8cdc4060fce075

harvested Assembly for continuous-domain attention mechanisms (graph method: continuous-domain attention mechanisms). 13 bricks; individually on a fixture: 12 ran (of which 9 constructed an object rather than computing a result; 9 with no instrument failure: 0 honoured, 0 violated, 9 with no contract checked; 3 where Syntology's instrument failed) · 1 unverified. The assembly has not been run end-to-end.

  1. ContinuousAttention entry · level 0 · unverified · 7bc1d8632d6b41c5
  2. ContinuousSoftmax model · level 1 · ran · 88d5d46a3e9f268f
  3. ContinuousSparsemax model · level 1 · ran · 91ec632f03a8f06b
  4. add_gaussian_basis_functions helper · level 1 · ran · our draft was wrong · b2b0552bda1b4c2d
  5. add_power_basis_functions helper · level 1 · ran · our draft was wrong · 4500f4cfd4afb4ee
  6. add_wave_basis_functions helper · level 1 · ran · our draft was wrong · 7811322c90f19d30
  7. ContinuousSoftmaxFunction helper · level 1 · ran · 2abd421a2409332c
  8. ContinuousSparsemaxFunction helper · level 1 · ran · 45b9c83a5ff5f057
  9. GaussianBasisFunctions helper · level 1 · ran · d4a4993b0f77f70a
  10. PowerBasisFunctions helper · level 1 · ran · af4a48bd715bb239
  11. CosineBasisFunctions helper · level 1 · ran · 0513653665d957f6
  12. SineBasisFunctions helper · level 1 · ran · a041cde2266ca5ff
  13. BasisFunctions helper · level 1 · ran · 544ace0143134f21

harvested Assembly for α-entmax extension to continuous domains (graph method: α-entmax extension to continuous domains). 2 bricks; individually on a fixture: 0 ran · 2 unverified. The assembly has not been run end-to-end.

  1. ContinuousSparsemax entry · level 0 · unverified · 4d564215130f97e9
  2. ContinuousSparsemaxFunction helper · level 0 · unverified · dc6bd3d88b59cbda

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="2006.07214") returns these assemblies with each brick's role and wiring (how to connect).

Tasks

Machine TranslationQuestion AnsweringText ClassificationTranslationVisual Question AnsweringVisual Question Answering (VQA)text-classification

Results from the paper archive 2025-07-28

TaskDatasetModelMetricValueRank at snapshotLeaderboardReport
Visual Question Answering (VQA) VQA v2 test-dev 2D continuous softmax Accuracy 65.96 #41 of 56 Archive leaderboard report
Visual Question Answering (VQA) VQA v2 test-std 2D continuous softmax overall 66.27 #35 of 38 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

SoftmaxSparsemax

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