Papers › T-SCEND: Test-time Scalable MCTS-enhanced Diffusion Model

T-SCEND: Test-time Scalable MCTS-enhanced Diffusion Model

4 Feb 2025arXiv:2502.01989archive 2025-07-28

Tao Zhang, Jia-Shu Pan, Ruiqi Feng, Tailin Wu

We introduce Test-time Scalable MCTS-enhanced Diffusion Model (T-SCEND), a novel framework that significantly improves diffusion model's reasoning capabilities with better energy-based training and scaling up test-time computation. We first show that naïvely scaling up inference budget for diffusion models yields marginal gain. To address this, the training of T-SCEND consists of a novel linear-regression negative contrastive learning objective to improve the performance-energy consistency of the energy landscape, and a KL regularization to reduce adversarial sampling. During inference, T-SCEND integrates the denoising process with a novel hybrid Monte Carlo Tree Search (hMCTS), which sequentially performs best-of-N random search and MCTS as denoising proceeds. On challenging reasoning tasks of Maze and Sudoku, we demonstrate the effectiveness of T-SCEND's training objective and scalable inference method. In particular, trained with Maze sizes of up to 6×6, our T-SCEND solves 88% of Maze problems with much larger sizes of 15×15, while standard diffusion completely fails.Code to reproduce the experiments can be found at https://github.com/AI4Science-WestlakeU/t_scend.

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.

Syntology holds a reference implementation of VFScale, 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='VFScale') (how to connect); the free check have('2502.01989') needs no token.

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

Code

Syntology 34 code samples harvested from 2 repositories linked to this paper: 26 ran (of which 0 constructed an object rather than computing a result; 19 with no instrument failure: 2 honoured, 3 violated, 14 with no contract checked; 7 where Syntology's instrument failed) · 8 unverified. Where Syntology's instrument failed: 5 where our contract draft was wrong; 2 where the contract check could not run; failures of Syntology's instrument, not of the code.

By repository: official repository: 13 samples from 1 repository, 12 ran (11 with no instrument failure; 1 where Syntology's instrument failed); found in paper text by Syntology: 21 samples from 1 repository, 14 ran (8 with no instrument failure; 6 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 between 2026-09-22 and 2026-09-25 (UTC): the dates Syntology's graph wrote each sample's verification level, not when the code executed. 5 of the 26 samples that ran have no recorded date.

ai4science-westlakeu/t_scend officialmentioned in papermentioned on GitHubpytorchMITSyntology: 13 samples, 12 ran (1 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

34 samples harvested: 26 ran (of which 0 constructed an object rather than computing a result; 19 with no instrument failure: 2 honoured, 3 violated, 14 with no contract checked; 7 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 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
5ran · our draft was wrong
2ran; the contract check could not run
14ran
8unverified

Licence: 0 of the 34 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”. “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.

ai4science-westlakeu/t_scend official repository 13 samples, 12 ran (1 instrument-failed) · 1 unverified
identity in 174 papers
census: 11 implementations from 39 papers, 3 distinct outputs across 2 buckets, one shared input per bucket this sample: bucket 1 of 2, arg 1: rank 2, kind float, dtype float32
tscend_src/model/diffusion_lib/denoising_diffusion_pytorch_1d.py ran · honoured contract MIT (permissive) · 7f1040f5e3991d5e · report fixture: signature t:Tensor/2/float, args:tuple/1/Tensor, kwargs:dict/0/? · returned Tensor · level recorded, date not recorded
swish in 258 papers
census: 14 implementations from 37 papers, 2 distinct outputs on one shared input this sample: bucket 1 of 1, arg 1: rank 2, kind float, dtype float32
tscend_src/model/models.py ran · our draft was wrong fingerprinted MIT (permissive) · 0f786c407fb1ee4c · report fixture: signature x:Tensor/2/float · returned Tensor · level recorded, date not recorded
add_args_from_config tscend_src/utils/utils.py ran MIT (permissive) · 120c5b54f80e01ab · report fixture: signature parser:ArgumentParser/0/? · returned ArgumentParser · level recorded 2026-09-24
broadcast in 21 papers tscend_src/model/diffusion_lib/nlm_utils.py ran MIT (permissive) · cf4d925ba3e0cd70 · report fixture: signature tensor:Tensor/2/float, dim:int/0/int, size:int/0/int · returned Tensor · level recorded 2026-09-25
caculate_confidence_interval tscend_src/utils/utils.py ran fingerprinted MIT (permissive) · d03ae14d2490cf24 · report fixture: signature data:Tensor/2/float · returned tuple · level recorded 2026-09-24
default in 430 papers tscend_src/model/diffusion_lib/denoising_diffusion_pytorch_1d.py ran · violated contract MIT (permissive) · d1ef6b8cb9a28a53 · report fixture: signature val:str/0/str, d:callable/0/? · returned str · 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
tscend_src/model/diffusion_lib/denoising_diffusion_pytorch_1d.py ran · violated contract MIT (permissive) · 608e364a9d2376a3 · report fixture: signature x:Tensor/1/int · returned bool · level recorded, date not recorded
get_time in 20 papers tscend_src/utils/utils.py ran MIT (permissive) · 35d5144a8f9cae87 · report fixture: signature is_bracket:bool/0/bool, return_numerical_time:bool/0/bool, precision:str/0/str · returned str · level recorded 2026-09-24
meshgrid in 8 papers tscend_src/model/diffusion_lib/nlm_utils.py ran MIT (permissive) · 3bd1f73dff951c9d · report fixture: signature input1:Tensor/3/float, input2:Tensor/3/float, dim:int/0/int · returned tuple · level recorded 2026-09-24
meshgrid_exclude_self tscend_src/model/diffusion_lib/nlm_utils.py ran MIT (permissive) · a8a55f19dcddf768 · report fixture: signature input:Tensor/4/float, dim:int/0/int · returned Tensor · level recorded 2026-09-24
process_tensor_sudoku tscend_src/utils/utils_maze_continuous.py ran MIT (permissive) · 465cda5a23bfa556 · report fixture: signature x:Tensor/4/float · returned Tensor · level recorded 2026-09-25
sudoku_score tscend_src/utils/utils_maze_continuous.py ran MIT (permissive) · bf38386f25f34c74 · report fixture: signature pred:Tensor/3/float · returned Tensor · level recorded 2026-09-24
AI4Science-WestlakeU/VFScale found in paper text by Syntology 21 samples, 14 ran (6 instrument-failed) · 7 unverified
Expander vfscale_src/model/models.py ran · metamorphic tier: invariant MIT (permissive) · 14f0772248b45045 · report fixture: signature input:Tensor/3/float, n:int/0/int · returned Tensor · level recorded 2026-09-22
InferenceBase vfscale_src/model/models.py ran MIT (permissive) · 143b2aeff2125292 · report fixture: signature inputs:Tensor/2/float · returned Tensor · level recorded 2026-09-24
LinearLayer in 14 papers vfscale_src/model/models.py ran · metamorphic tier: invariant MIT (permissive) · 8d87e276de13f530 · report fixture: signature input:Tensor/2/float · returned Tensor · level recorded 2026-09-22
LogicInference vfscale_src/model/models.py ran MIT (permissive) · b20179518aab1891 · report fixture: signature inputs:Tensor/3/float · returned Tensor · level recorded 2026-09-24
MLPLayer in 15 papers vfscale_src/model/models.py ran MIT (permissive) · b172f5e1fa3d4d6a · report fixture: signature input:Tensor/2/float · returned Tensor · level recorded 2026-09-24
Reducer vfscale_src/model/models.py ran · metamorphic tier: invariant MIT (permissive) · 4d7e91ea410a99e5 · report fixture: signature input:Tensor/4/float · returned Tensor · level recorded 2026-09-22
_get_tuple_n vfscale_src/model/models.py ran · honoured contract MIT (permissive) · 421df320f85fd119 · report fixture: signature x:int/0/int, n:int/0/int, tp:callable/0/? · returned list · level recorded 2026-09-22
broadcast in 21 papers vfscale_src/model/models.py ran; the contract check could not run (RAISES) MIT (permissive) · a58996a6b5a75e87 · report fixture: signature tensor:Tensor/2/float, dim:int/0/int, size:int/0/int · returned Tensor · level recorded 2026-09-22
exclude_mask vfscale_src/model/models.py ran; the contract check could not run (RAISES) MIT (permissive) · 70af23bab509259e · report fixture: signature input:Tensor/4/float, cnt:int/0/int, dim:int/0/int · returned Tensor · level recorded 2026-09-22
get_activation in 218 papers vfscale_src/model/models.py ran · our draft was wrong MIT (permissive) · c608fc7d98f990e0 · report fixture: signature act:str/0/str · returned ReLU · level recorded 2026-09-22
get_batcnnorm vfscale_src/model/models.py ran · our draft was wrong MIT (permissive) · 9b9856f78988b052 · report fixture: signature bn:callable/0/?, nr_features:int/0/int, nr_dims:int/0/int · returned BatchNorm1d · level recorded 2026-09-22
get_dropout vfscale_src/model/models.py ran · our draft was wrong MIT (permissive) · 3acb2b23a9bd22b9 · report fixture: signature dropout:float/0/float, nr_dims:int/0/int · returned Dropout · level recorded 2026-09-22
mask_value vfscale_src/model/models.py ran · our draft was wrong fingerprinted MIT (permissive) · dbb922152c1e332c · report fixture: signature input:Tensor/2/float, mask:Tensor/2/float, value:float/0/float · returned Tensor · level recorded 2026-09-22
str2bool in 533 papers vfscale_src/inference/inference_Maze.py ran · violated contract MIT (permissive) · 7c54b06017e05bd2 · report fixture: signature x:str/0/str · returned bool · level recorded, date not recorded
GNNConvEBM vfscale_src/model/models.py unverified MIT (permissive) · 36f1484dd68e104d · report
LogicLayer vfscale_src/model/models.py unverified MIT (permissive) · a5425637206e3216 · report
LogicMachine vfscale_src/model/models.py unverified MIT (permissive) · 6615bbe76e18ddd6 · report
NLMConv1DBlock vfscale_src/model/models.py unverified MIT (permissive) · 2221e5ee9ad31dc0 · report
NLMConvBlock vfscale_src/model/models.py unverified MIT (permissive) · 80023fe7520ff23f · report
Permutation in 5 papers vfscale_src/model/models.py unverified MIT (permissive) · bbedff446942642a · report
concat_shape vfscale_src/model/models.py unverified MIT (permissive) · b40432e68b551baa · 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 VFScale (graph method: VFScale), 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 3; 5 property tests passed on both).

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

  1. ucb_score entry · level 3 · ran · 5b23b3392a7ffb3274f9da61dc248cbc2b436361acc60818b8b45aef36c220bf

harvested Assembly for VFScale (graph method: VFScale). 22 bricks; individually on a fixture: 15 ran (of which 1 constructed an object rather than computing a result; 8 with no instrument failure: 1 honoured, 0 violated, 7 with no contract checked; 7 where Syntology's instrument failed) · 7 unverified. The assembly has not been run end-to-end.

  1. GNNConvEBM entry · level 0 · unverified · 36f1484dd68e104d
  2. NLMConv1DBlock helper · level 0 · unverified · 2221e5ee9ad31dc0
  3. NLMConvBlock model · level 0 · unverified · 80023fe7520ff23f
  4. SinusoidalPosEmb helper · level 1 · ran · 3f9e9512b765d426
  5. LogicMachine model · level 0 · unverified · 6615bbe76e18ddd6
  6. swish helper · level 2 · ran · our draft was wrong · 0f786c407fb1ee4c
  7. LogicLayer helper · level 0 · unverified · a5425637206e3216
  8. Expander helper · level 1 · ran · 14f0772248b45045
  9. LogicInference helper · level 1 · ran · b20179518aab1891
  10. Permutation helper · level 0 · unverified · bbedff446942642a
  11. Reducer helper · level 1 · ran · 4d7e91ea410a99e5
  12. _get_tuple_n helper · level 1 · ran · honoured contract · 421df320f85fd119
  13. broadcast helper · level 1 · ran; the contract check could not run (RAISES) · a58996a6b5a75e87
  14. InferenceBase helper · level 1 · ran · 143b2aeff2125292
  15. exclude_mask helper · level 1 · ran; the contract check could not run (RAISES) · 70af23bab509259e
  16. mask_value helper · level 1 · ran · our draft was wrong · dbb922152c1e332c
  17. concat_shape helper · level 0 · unverified · b40432e68b551baa
  18. MLPLayer helper · level 1 · ran · b172f5e1fa3d4d6a
  19. LinearLayer helper · level 1 · ran · 8d87e276de13f530
  20. get_activation helper · level 1 · ran · our draft was wrong · c608fc7d98f990e0
  21. get_batcnnorm helper · level 1 · ran · our draft was wrong · 9b9856f78988b052
  22. get_dropout helper · level 1 · ran · our draft was wrong · 3acb2b23a9bd22b9

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="2502.01989") returns these assemblies with each brick's role and wiring (how to connect).

Tasks

Contrastive LearningDenoising

Results from the paper archive 2025-07-28

No leaderboard rows for this paper in the archive.

Methods

Contrastive LearningDiffusionRandom Search

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