Papers › Actions Speak Louder than Words: Trillion-Parameter Sequential Transducers for...
Actions Speak Louder than Words: Trillion-Parameter Sequential Transducers for Generative Recommendations
Jiaqi Zhai, Lucy Liao, Xing Liu, Yueming Wang, Rui Li, Xuan Cao, Leon Gao, Zhaojie Gong, Fangda Gu, Michael He, Yinghai Lu, Yu Shi
Large-scale recommendation systems are characterized by their reliance on high cardinality, heterogeneous features and the need to handle tens of billions of user actions on a daily basis. Despite being trained on huge volume of data with thousands of features, most Deep Learning Recommendation Models (DLRMs) in industry fail to scale with compute. Inspired by success achieved by Transformers in language and vision domains, we revisit fundamental design choices in recommendation systems. We reformulate recommendation problems as sequential transduction tasks within a generative modeling framework ("Generative Recommenders"), and propose a new architecture, HSTU, designed for high cardinality, non-stationary streaming recommendation data. HSTU outperforms baselines over synthetic and public datasets by up to 65.8% in NDCG, and is 5.3x to 15.2x faster than FlashAttention2-based Transformers on 8192 length sequences. HSTU-based Generative Recommenders, with 1.5 trillion parameters, improve metrics in online A/B tests by 12.4% and have been deployed on multiple surfaces of a large internet platform with billions of users. More importantly, the model quality of Generative Recommenders empirically scales as a power-law of training compute across three orders of magnitude, up to GPT-3/LLaMa-2 scale, which reduces carbon footprint needed for future model developments, and further paves the way for the first foundational models in recommendations.
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="2402.17152")get_code(code_sha256="…")with a sample'scode_sha256from the list belowget_citation_path(paper_1="2402.17152", paper_2="…")with another paper's arXiv id or titleget_concepts_for_paper(arxiv_id="2402.17152")
Code
Syntology 42 code samples harvested from 8 repositories linked to this paper: 15 ran (of which 13 constructed an object rather than computing a result; 14 with no instrument failure: 0 honoured, 0 violated, 14 with no contract checked; 1 where Syntology's instrument failed) · 27 unverified. Where Syntology's instrument failed: 1 where the contract check could not run; failures of Syntology's instrument, not of the code.
By repository: official repository: 9 samples from 1 repository, 2 ran (2 with no instrument failure; 0 where Syntology's instrument failed); community (archive-listed): 33 samples from 7 repositories, 13 ran (12 with no instrument failure; 1 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 2 of its repositories that Syntology harvested for other papers:
- MindSpore-scientific/code-12, community (archive-listed): harvested for another paper, listed under the repository
- pwc-1/Paper-9, 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.
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
42 samples harvested: 15 ran (of which 13 constructed an object rather than computing a result; 14 with no instrument failure: 0 honoured, 0 violated, 14 with no contract checked; 1 where Syntology's instrument failed) · 27 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: 3 of the 42 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 8 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 4 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.
87d520a2cecef5ab · report
fixture: returned RelativeBucketedTimeAndPositionBasedBias · level recorded 2026-09-22 8e67290b484621dc · report
fixture: returned SequentialTransductionUnitJagged · level recorded 2026-09-22 1b0e681749e03d79 · report
af3ba853909518b7 · report
69baf41bf38fa53a · report
722d4d3551b1330b · report
c7bd477551eea3d8 · report
5cf1b1643aeb298f · report
e4eacfaee3c9dde6 · report
85a6849359c5f5e8 · report
fixture: returned RankedLogger · level recorded 2026-09-22 45a12d3d33925880 · report
fixture: returned RelativeBucketedTimeAndPositionBasedBias · level recorded 2026-09-22 da2a3d953e663a05 · report
fixture: returned SequentialTransductionUnitJagged · level recorded 2026-09-22 9b97ef3be03bed65 · report
9cff21d1111d44eb · report
7eaf0c9a3f641a2a · report
732a9292a549f889 · report
fixture: returned GeneralizedInteractionModule · level recorded 2026-09-22 cf69abc69c9a52b1 · report
fixture: returned RelativeBucketedTimeAndPositionBasedBias · level recorded 2026-09-22 f1c140f6503605bf · report
fixture: returned SequentialTransductionUnitJagged · level recorded 2026-09-22 0376ea3757f5f63e · report
fixture: signature lengths:Tensor/1/int, encoded_embeddings:Tensor/3/float · returned Tensor · level recorded 2026-09-22 eece677a59828947 · report
724ed6e7e00a7fda · report
501db889bfdd9bb0 · report
9269aeddc68cac36 · report
3579714787dc8d39 · report
b0c190ff62ae0dd3 · report
f19becb1418ef093 · report
514d4d6606c0e9ad · report
07e6e09a56bb2aab · report
fixture: signature a:Tensor/2/float, b:Tensor/2/float · returned tuple · level recorded 2026-09-24 21d9bc0ad504318b · report
d3b2601e4e7b9ead · report
1d89041488291bb5 · report
b7537cbdee06dac7 · report
fixture: returned SequentialEncoderWithLearnedSimilarityModule · level recorded 2026-09-22 3c1986d68d087d2a · report
fixture: returned SequentialTransductionUnitJagged · level recorded 2026-09-22 a67884cc50d40aad · report
41be69705279a075 · report
d5bd879a7af1a427 · report
a3d0818722016cf5 · report
d710bcff6a3d3d81 · report
fixture: returned HSTUBlock · level recorded 2026-09-22 2e940ea8095c3ba3 · report
fixture: returned PointwiseAggregatedAttention · level recorded 2026-09-22 c0d8162f6ccbc40a · report
fixture: returned RelativeAttentionBias · level recorded 2026-09-22 543debc1b00b95a3 · 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. 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 HSTU (harvested), lifted in the same batch on 2026-09-09: alternative readings kept side by side; none replaces or supersedes another.
harvested Assembly for HSTU (graph method: HSTU). 12 bricks; individually on a fixture: 4 ran (of which 3 constructed an object rather than computing a result; 3 with no instrument failure: 0 honoured, 0 violated, 3 with no contract checked; 1 where Syntology's instrument failed) · 8 unverified. The assembly has not been run end-to-end.
one of 7 siblings
- HSTU entry · level 0 · unverified ·
724ed6e7e00a7fda - EmbeddingModule helper · level 0 · unverified ·
eece677a59828947 - GeneralizedInteractionModule helper · level 1 · ran ·
732a9292a549f889 - HSTUJagged model · level 0 · unverified ·
501db889bfdd9bb0 - InputFeaturesPreprocessorModule helper · level 0 · unverified ·
9269aeddc68cac36 - NDPModule helper · level 0 · unverified ·
b0c190ff62ae0dd3 - OutputPostprocessorModule helper · level 0 · unverified ·
f19becb1418ef093 - RelativeBucketedTimeAndPositionBasedBias helper · level 1 · ran ·
cf69abc69c9a52b1 - SequentialTransductionUnitJagged model · level 1 · ran ·
f1c140f6503605bf - get_current_embeddings helper · level 1 · ran; the contract check could not run (RAISES) ·
0376ea3757f5f63e - InteractionModule helper · level 0 · unverified ·
3579714787dc8d39 - RelativeAttentionBiasModule helper · level 0 · unverified ·
514d4d6606c0e9ad
harvested Assembly for HSTU (graph method: HSTU). 12 bricks; individually on a fixture: 3 ran (of which 2 constructed an object rather than computing a result; 2 with no instrument failure: 0 honoured, 0 violated, 2 with no contract checked; 1 where Syntology's instrument failed) · 9 unverified. The assembly has not been run end-to-end.
one of 7 siblings
- HSTU entry · level 0 · unverified ·
1b0e681749e03d79 - EmbeddingModule helper · level 0 · unverified ·
eece677a59828947 - HSTUJagged helper · level 0 · unverified ·
af3ba853909518b7 - InputFeaturesPreprocessorModule helper · level 0 · unverified ·
69baf41bf38fa53a - OutputPostprocessorModule helper · level 0 · unverified ·
f19becb1418ef093 - RelativeBucketedTimeAndPositionBasedBias model · level 1 · ran ·
87d520a2cecef5ab - SequentialEncoderWithLearnedSimilarityModule helper · level 0 · unverified ·
c7bd477551eea3d8 - SequentialTransductionUnitJagged model · level 1 · ran ·
8e67290b484621dc - SimilarityModule helper · level 0 · unverified ·
5cf1b1643aeb298f - get_current_embeddings helper · level 1 · ran; the contract check could not run (RAISES) ·
0376ea3757f5f63e - RelativeAttentionBiasModule helper · level 0 · unverified ·
722d4d3551b1330b - _hstu_attention_maybe_from_cache helper · level 0 · unverified ·
e4eacfaee3c9dde6
harvested Assembly for HSTU (graph method: HSTU). 7 bricks; individually on a fixture: 3 ran (of which 3 constructed an object rather than computing a result; 3 with no instrument failure: 0 honoured, 0 violated, 3 with no contract checked; 0 where Syntology's instrument failed) · 4 unverified; every one of the 3 samples that ran constructed an object rather than computing a result. The assembly has not been run end-to-end.
one of 7 siblings
- HSTU entry · level 0 · unverified ·
9b97ef3be03bed65 - HSTUJagged model · level 0 · unverified ·
9cff21d1111d44eb - RelativeBucketedTimeAndPositionBasedBias model · level 1 · ran ·
45a12d3d33925880 - SequentialTransductionUnitJagged model · level 1 · ran ·
da2a3d953e663a05 - RelativeAttentionBiasModule helper · level 0 · unverified ·
722d4d3551b1330b - _hstu_attention_maybe_from_cache helper · level 0 · unverified ·
7eaf0c9a3f641a2a - RankedLogger helper · level 1 · ran ·
85a6849359c5f5e8
harvested Assembly for HSTU (graph method: HSTU). 12 bricks; individually on a fixture: 4 ran (of which 3 constructed an object rather than computing a result; 3 with no instrument failure: 0 honoured, 0 violated, 3 with no contract checked; 1 where Syntology's instrument failed) · 8 unverified. The assembly has not been run end-to-end.
one of 7 siblings
- HSTU entry · level 0 · unverified ·
5f35558580c55d6d - EmbeddingModule helper · level 0 · unverified ·
eece677a59828947 - HSTUJagged model · level 0 · unverified ·
a67884cc50d40aad - InputFeaturesPreprocessorModule helper · level 0 · unverified ·
41be69705279a075 - OutputPostprocessorModule helper · level 0 · unverified ·
f19becb1418ef093 - RelativeBucketedTimeAndPositionBasedBias helper · level 1 · ran ·
87d520a2cecef5ab - SequentialEncoderWithLearnedSimilarityModule helper · level 1 · ran ·
b7537cbdee06dac7 - SequentialTransductionUnitJagged model · level 1 · ran ·
3c1986d68d087d2a - SimilarityModule helper · level 0 · unverified ·
d5bd879a7af1a427 - get_current_embeddings helper · level 1 · ran; the contract check could not run (RAISES) ·
0376ea3757f5f63e - RelativeAttentionBiasModule helper · level 0 · unverified ·
722d4d3551b1330b - _hstu_attention_maybe_from_cache model · level 0 · unverified ·
a3d0818722016cf5
harvested Assembly for HSTU (graph method: HSTU). 3 bricks; individually on a fixture: 3 ran (of which 3 constructed an object rather than computing a result; 3 with no instrument failure: 0 honoured, 0 violated, 3 with no contract checked; 0 where Syntology's instrument failed) · 0 unverified; every one of the 3 samples that ran constructed an object rather than computing a result. The assembly has not been run end-to-end.
one of 7 siblings
- HSTUBlock entry · level 1 · ran ·
d710bcff6a3d3d81 - PointwiseAggregatedAttention model · level 1 · ran ·
2e940ea8095c3ba3 - RelativeAttentionBias model · level 1 · ran ·
c0d8162f6ccbc40a
harvested Assembly for HSTU (graph method: HSTU). 12 bricks; individually on a fixture: 3 ran (of which 2 constructed an object rather than computing a result; 2 with no instrument failure: 0 honoured, 0 violated, 2 with no contract checked; 1 where Syntology's instrument failed) · 9 unverified. The assembly has not been run end-to-end.
one of 7 siblings
- HSTU entry · level 0 · unverified ·
543debc1b00b95a3 - EmbeddingModule helper · level 0 · unverified ·
eece677a59828947 - HSTUJagged model · level 0 · unverified ·
af3ba853909518b7 - InputFeaturesPreprocessorModule helper · level 0 · unverified ·
69baf41bf38fa53a - OutputPostprocessorModule helper · level 0 · unverified ·
f19becb1418ef093 - RelativeBucketedTimeAndPositionBasedBias helper · level 1 · ran ·
87d520a2cecef5ab - SequentialEncoderWithLearnedSimilarityModule helper · level 0 · unverified ·
c7bd477551eea3d8 - SequentialTransductionUnitJagged model · level 1 · ran ·
8e67290b484621dc - SimilarityModule helper · level 0 · unverified ·
5cf1b1643aeb298f - get_current_embeddings helper · level 1 · ran; the contract check could not run (RAISES) ·
0376ea3757f5f63e - RelativeAttentionBiasModule helper · level 0 · unverified ·
722d4d3551b1330b - _hstu_attention_maybe_from_cache model · level 0 · unverified ·
e4eacfaee3c9dde6
harvested Assembly for HSTU (graph method: HSTU). 3 bricks; individually on a fixture: 0 ran · 3 unverified. The assembly has not been run end-to-end.
one of 7 siblings
- SequentialTransductionUnitJagged entry · level 0 · unverified ·
d3b2601e4e7b9ead - RelativeAttentionBiasModule helper · level 0 · unverified ·
21d9bc0ad504318b - _hstu_attention_maybe_from_cache helper · level 0 · unverified ·
1d89041488291bb5
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="2402.17152") 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 |
|---|---|---|---|---|---|---|---|
| Recommendation Systems | Amazon-Book | HSTU | HR@10 | 0.0478 | #15 of 16 | Archive leaderboard | report |
| Recommendation Systems | Amazon-Book | HSTU | HR@50 | 0.1082 | #15 of 16 | Archive leaderboard | report |
| Recommendation Systems | Amazon-Book | HSTU | NDCG@10 | 0.0262 | #15 of 16 | Archive leaderboard | report |
| Recommendation Systems | Amazon-Book | HSTU | NDCG@50 | 0.0393 | #15 of 16 | Archive leaderboard | report |
| Recommendation Systems | MovieLens 1M | HSTU | HR@10 (full corpus) | 0.3294 | #29 of 31 | Archive leaderboard | report |
| Recommendation Systems | MovieLens 1M | HSTU | NDCG@10 (full corpus) | 0.1893 | #29 of 31 | Archive leaderboard | report |
| Recommendation Systems | MovieLens 20M | HSTU | HR@10 (full corpus) | 0.3556 | #15 of 18 | Archive leaderboard | report |
| Recommendation Systems | MovieLens 20M | HSTU | nDCG@10 (full corpus) | 0.2098 | #15 of 18 | 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.
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