Papers › Self-Attention Between Datapoints: Going Beyond Individual Input-Output Pairs in Deep Learning

Self-Attention Between Datapoints: Going Beyond Individual Input-Output Pairs in Deep Learning

4 Jun 2021NeurIPS 2021 12arXiv:2106.02584archive 2025-07-28

Jannik Kossen, Neil Band, Clare Lyle, Aidan N. Gomez, Tom Rainforth, Yarin Gal

We challenge a common assumption underlying most supervised deep learning: that a model makes a prediction depending only on its parameters and the features of a single input. To this end, we introduce a general-purpose deep learning architecture that takes as input the entire dataset instead of processing one datapoint at a time. Our approach uses self-attention to reason about relationships between datapoints explicitly, which can be seen as realizing non-parametric models using parametric attention mechanisms. However, unlike conventional non-parametric models, we let the model learn end-to-end from the data how to make use of other datapoints for prediction. Empirically, our models solve cross-datapoint lookup and complex reasoning tasks unsolvable by traditional deep learning models. We show highly competitive results on tabular data, early results on CIFAR-10, and give insight into how the model makes use of the interactions between points.

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 self-attention based deep learning architecture, generated from the paper text (not the authors' code), at level 3 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)”; V3 “paper-derived property tests pass on both of those same two generations.” (quoted from Syntology's server card). No rung compares this code against the authors' implementation. For agents: get_reference_implementation(name='self-attention based deep learning architecture') (how to connect); the free check have('2106.02584') needs no token.

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

Code

Syntology 30 code samples harvested from 2 repositories linked to this paper: 22 ran (of which 4 constructed an object rather than computing a result; 20 with no instrument failure: 2 honoured, 1 violated, 17 with no contract checked; 2 where Syntology's instrument failed) · 8 unverified. Where Syntology's instrument failed: 2 where our contract draft was wrong; failures of Syntology's instrument, not of the code.

By repository: official repository: 11 samples from 1 repository, 6 ran (6 with no instrument failure; 0 where Syntology's instrument failed); community (archive-listed): 19 samples from 1 repository, 16 ran (14 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.

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. 3 of the 22 samples that ran have no recorded date.

OATML/Non-Parametric-Transformers officialmentioned in papermentioned on GitHubpytorchSyntology: 11 samples, 6 ran (0 instrument-failed) · 5 unverified report
oatml-markslab/proteinnpt mentioned on GitHubpytorchMITSyntology: 19 samples, 16 ran (2 instrument-failed) · 3 unverified report
richrast/spin mentioned on GitHubpytorchApache-2.0Syntology: no sample linked to this paper (harvested for another paper). Syntology's graph links these samples to Semi-Parametric Inducing Point Networks and Neural Processes (arXiv:2205.11718; that paper's own run record: 10 ran (of which 4 constructed an object rather than computing a result; 10 with no instrument failure: 0 honoured, 0 violated, 10 with no contract checked; 0 where Syntology's instrument failed) · 5 unverified); nothing checks that they implement this paper's method. 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

30 samples harvested: 22 ran (of which 4 constructed an object rather than computing a result; 20 with no instrument failure: 2 honoured, 1 violated, 17 with no contract checked; 2 where Syntology's instrument failed) · 8 unverified. “Where Syntology's instrument failed” counts failures of Syntology's instrument, not of the code. Syntology's record for this page last changed ; when this build read Syntology's graph is in the build record. That date is not when the samples ran.

2ran · honoured contract
1ran · violated contract
2ran · our draft was wrong
17ran
8unverified

Licence: 0 of the 30 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 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.

OATML/Non-Parametric-Transformers official repository 11 samples, 6 ran (0 instrument-failed) · 5 unverified
count_parameters in 278 papers npt/utils/train_utils.py ran · honoured contract Apache-2.0 (permissive) · 0e18add484f1a90c · report fixture: signature model:callable/0/? · returned int · level recorded, date not recorded
Args in 6 papers npt/model/npt.py ran Apache-2.0 (permissive) · a94fc028457ce3f8 · report fixture: returned Args · level recorded 2026-09-22
MAB in 12 papers npt/model/npt.py ran · metamorphic tier: deterministic Apache-2.0 (permissive) · 7232573389ce2a5a · report fixture: returned MAB · level recorded 2026-09-22
Print npt/model/npt.py ran · metamorphic tier: deterministic fingerprinted Apache-2.0 (permissive) · d6f128d3e349e966 · report fixture: returned Print · level recorded 2026-09-22
ReshapeToFlat npt/model/npt.py ran · metamorphic tier: invariant fingerprinted Apache-2.0 (permissive) · 0ead572d1093cc14 · report fixture: signature X:Tensor/3/float · returned Tensor · level recorded 2026-09-22
ReshapeToNested npt/model/npt.py ran · metamorphic tier: deterministic fingerprinted Apache-2.0 (permissive) · e6ffb4964d98205f · report fixture: returned ReshapeToNested · level recorded 2026-09-22
ImagePatcher npt/model/npt.py unverified Apache-2.0 (permissive) · 8d76f18f88c52502 · report
LinearImagePatcher npt/model/npt.py unverified Apache-2.0 (permissive) · ce48ecd197be4b48 · report
MHSA npt/model/npt.py unverified Apache-2.0 (permissive) · f3dd2c06a9c7eb93 · report
NPTModel npt/model/npt.py unverified Apache-2.0 (permissive) · 611506804344b302 · report
SaveAttMaps npt/model/npt.py unverified Apache-2.0 (permissive) · 67c82e2cb5a619f5 · report
oatml-markslab/proteinnpt community (archive-listed) 19 samples, 16 ran (2 instrument-failed) · 3 unverified
PNPT_sample_training_points_inference proteinnpt/proteinnpt/data_processing.py ran MIT (permissive) · d2d1d1bc5e8b6b1d · report fixture: signature training_sequences:dict/0/?, sequences_sampling_method:str/0/str, num_sampled_points:int/0/int · returned ndarray · level recorded 2026-09-24
apc in 6 papers proteinnpt/utils/esm/modules.py ran · our draft was wrong fingerprinted MIT (permissive) · f6701ae421cbe0d4 · report fixture: signature x:Tensor/4/float · returned Tensor · level recorded 2026-09-24
assign_group_stratified proteinnpt/utils/cv_split.py ran fingerprinted MIT (permissive) · 7fc8e20691370cda · report fixture: signature x:ndarray/1/int · returned ndarray · level recorded 2026-09-24
compute_weight in 8 papers proteinnpt/utils/msa_utils.py ran MIT (permissive) · 7b4449eb96bfa4eb · report fixture: signature seq:ndarray/1/float, list_seq:ndarray/2/float, theta:float/0/float · returned float · level recorded 2026-09-24
compute_weight_all proteinnpt/utils/msa_utils.py ran MIT (permissive) · 0d397ce1b8012be1 · report fixture: signature list_seq:ndarray/2/float, theta:float/0/float, updated_freq:int/0/int · returned ndarray · level recorded 2026-09-24
gelu in 308 papers
census: 43 implementations from 74 papers, 6 distinct outputs on one shared input
proteinnpt/utils/esm/modules.py ran · our draft was wrong fingerprinted MIT (permissive) · 2e7bd6c4ccd1ed68 · report fixture: signature x:Tensor/2/float · returned Tensor · level recorded 2026-09-24
get_learning_rate in 19 papers proteinnpt/utils/model_utils.py ran MIT (permissive) · 7dabe8f85e5f4e16 · report fixture: signature training_step:int/0/int, num_warmup_steps:int/0/int, num_total_training_steps:int/0/int, max_learning_rate:float/0/float, min_learning_rate:float/0/float · returned float · level recorded 2026-09-24
get_parameter_names in 7 papers proteinnpt/utils/model_utils.py ran MIT (permissive) · a8991266f955dd35 · report fixture: signature model:callable/0/?, forbidden_layer_types:list/1/callable · returned list · level recorded 2026-09-24
get_slopes in 6 papers proteinnpt/utils/tranception/model_pytorch.py ran · honoured contract fingerprinted MIT (permissive) · 4305f4fdd8e54bc1 · report fixture: signature n:int/0/int, mode:str/0/str, verbose:bool/0/bool · returned list · level recorded, date not recorded
keep_singles proteinnpt/utils/cv_split.py ran MIT (permissive) · ea7a41d551b7a7e5 · report fixture: signature DMS:DataFrame/0/?, mutant_column:str/0/str · returned DataFrame · level recorded 2026-09-24
learning_rate_scheduler proteinnpt/utils/model_utils.py ran MIT (permissive) · 355bff3ffb0d66a7 · report fixture: signature num_warmup_steps:int/0/int, num_total_training_steps:int/0/int, max_learning_rate:float/0/float, min_learning_rate:float/0/float · returned function · level recorded 2026-09-24
one_hot_3D proteinnpt/utils/msa_utils.py ran MIT (permissive) · 52a00d097b2f46fd · report fixture: signature seq_keys:list/1/str, seq_name_to_sequence:dict/0/?, alphabet:str/0/str, seq_length:int/0/int · returned ndarray · level recorded 2026-09-24
return_position_single proteinnpt/utils/cv_split.py ran MIT (permissive) · 4074372860cf12f3 · report fixture: signature mutation:str/0/str · returned int · level recorded 2026-09-24
split_data_based_on_test_fold_index proteinnpt/utils/data_utils.py ran MIT (permissive) · 72ad19ee550806af · report fixture: signature dataframe:DataFrame/0/?, fold_variable_name:str/0/str, test_fold_index:int/0/int, use_validation_set:bool/0/bool · returned tuple · level recorded 2026-09-24
standardize in 47 papers proteinnpt/utils/data_utils.py ran fingerprinted MIT (permissive) · 74862b2cc41d3a95 · report fixture: signature x:ndarray/2/float, epsilon:float/0/float · returned ndarray · level recorded 2026-09-24
symmetrize in 19 papers proteinnpt/utils/esm/modules.py ran · violated contract fingerprinted MIT (permissive) · b385be9826c4a8b3 · report fixture: signature x:Tensor/3/float · returned Tensor · level recorded, date not recorded
get_tranception_tokenizer proteinnpt/utils/tranception/model_pytorch.py unverified MIT (permissive) · e20cd5502a6c2414 · report
nanmean in 9 papers proteinnpt/utils/tranception/model_pytorch.py unverified MIT (permissive) · 5574012ba4af04ff · report
process_batch in 31 papers proteinnpt/baselines/data_processing.py unverified MIT (permissive) · c5e4288eb459c27d · 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.

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

generated Assembly for self-attention based deep learning architecture (graph method: self-attention based deep learning architecture), generated from the paper text, not the repository's code. 1 brick; individually on a fixture: 1 ran (1 with no instrument failure: 1 honoured, 0 violated, 0 with no contract checked; 0 where Syntology's instrument failed) · 0 unverified. Ran end-to-end against an independent peer (level 3; 5 property tests passed on both).

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

  1. compute_relative_prediction_change entry · level 3 · ran · honoured contract · 27eb9508b5bcd234d5e48e82902d7b8d34aefe5c2776f0e3462b327487d4a0b0

harvested Assembly for self-attention based deep learning architecture (graph method: self-attention based deep learning architecture). 11 bricks; individually on a fixture: 6 ran (of which 4 constructed an object rather than computing a result; 5 with no instrument failure: 0 honoured, 0 violated, 5 with no contract checked; 1 where Syntology's instrument failed) · 5 unverified. The assembly has not been run end-to-end.

  1. NPTModel entry · level 0 · unverified · 611506804344b302
  2. Args helper · level 1 · ran · a94fc028457ce3f8
  3. MHSA model · level 0 · unverified · f3dd2c06a9c7eb93
  4. Print helper · level 1 · ran · d6f128d3e349e966
  5. ReshapeToFlat helper · level 1 · ran · 0ead572d1093cc14
  6. ReshapeToNested helper · level 1 · ran · e6ffb4964d98205f
  7. torch_cast_to_dtype helper · level 1 · ran; the contract check could not run · c3dd55e49fd28138
  8. MAB model · level 1 · ran · 7232573389ce2a5a
  9. SaveAttMaps helper · level 0 · unverified · 67c82e2cb5a619f5
  10. ImagePatcher helper · level 0 · unverified · 8d76f18f88c52502
  11. LinearImagePatcher helper · level 0 · unverified · ce48ecd197be4b48

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="2106.02584") returns these assemblies with each brick's role and wiring (how to connect).

Tasks

3D Part SegmentationDeep Learning

Results from the paper archive 2025-07-28

No leaderboard rows for this paper in the archive.

Methods

1cycleAdaptive MaskingAdaptive Span TransformerAttentionAttention DropoutDense ConnectionsDropoutEmbedding DropoutL1 RegularizationLayer NormalizationLinear LayerMulti-Head AttentionReLUResidual ConnectionSoftmax

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