Papers › Flow Network based Generative Models for Non-Iterative Diverse Candidate Generation

Flow Network based Generative Models for Non-Iterative Diverse Candidate Generation

8 Jun 2021NeurIPS 2021 12arXiv:2106.04399archive 2025-07-28

Emmanuel Bengio, Moksh Jain, Maksym Korablyov, Doina Precup, Yoshua Bengio

This paper is about the problem of learning a stochastic policy for generating an object (like a molecular graph) from a sequence of actions, such that the probability of generating an object is proportional to a given positive reward for that object. Whereas standard return maximization tends to converge to a single return-maximizing sequence, there are cases where we would like to sample a diverse set of high-return solutions. These arise, for example, in black-box function optimization when few rounds are possible, each with large batches of queries, where the batches should be diverse, e.g., in the design of new molecules. One can also see this as a problem of approximately converting an energy function to a generative distribution. While MCMC methods can achieve that, they are expensive and generally only perform local exploration. Instead, training a generative policy amortizes the cost of search during training and yields to fast generation. Using insights from Temporal Difference learning, we propose GFlowNet, based on a view of the generative process as a flow network, making it possible to handle the tricky case where different trajectories can yield the same final state, e.g., there are many ways to sequentially add atoms to generate some molecular graph. We cast the set of trajectories as a flow and convert the flow consistency equations into a learning objective, akin to the casting of the Bellman equations into Temporal Difference methods. We prove that any global minimum of the proposed objectives yields a policy which samples from the desired distribution, and demonstrate the improved performance and diversity of GFlowNet on a simple domain where there are many modes to the reward function, and on a molecule synthesis task.

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 GFlowNet, 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='GFlowNet') (how to connect); the free check have('2106.04399') needs no token.

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

Code

Syntology 31 code samples: 28 from 6 repositories linked to this paper, 3 identical to code harvested from another paper's repository. Across all 31: 25 ran (of which 2 constructed an object rather than computing a result; 15 with no instrument failure: 6 honoured, 1 violated, 8 with no contract checked; 10 where Syntology's instrument failed) · 6 unverified. Where Syntology's instrument failed: 8 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: 15 samples from 1 repository, 12 ran (11 with no instrument failure; 1 where Syntology's instrument failed); community (archive-listed): 12 samples from 4 repositories, 9 ran (4 with no instrument failure; 5 where Syntology's instrument failed); community: 1 sample from 1 repository, 1 ran (0 with no instrument failure; 1 where Syntology's instrument failed); 3 identical to code first harvested elsewhere. 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 2026-09-26 (UTC): the dates Syntology's graph wrote each sample's verification level, not when the code executed. 12 of the 25 samples that ran have no recorded date.

bengioe/gflownet officialmentioned in papermentioned on GitHubpytorchMITSyntology: 15 samples, 12 ran (1 instrument-failed) · 3 unverified report
SeonghwanSeo/RxnFlow mentioned on GitHubpytorchSyntology: 1 sample, 1 ran (0 instrument-failed) · 0 unverified report
gfnorg/gflownet mentioned on GitHubpytorchSyntology: 5 samples, 3 ran (2 instrument-failed) · 2 unverified report
lazaratan/gflownet-generalization mentioned on GitHubpytorchSyntology: no sample linked to this paper (harvested for another paper). Syntology's graph links these samples to Learning GFlowNets from partial episodes for improved convergence and stability (arXiv:2209.12782; that paper's own run record: 6 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; 4 where Syntology's instrument failed) · 1 unverified); nothing checks that they implement this paper's method. report
princetonlips/mam mentioned on GitHubpytorchSyntology: 3 samples, 3 ran (1 instrument-failed) · 0 unverified report
recursionpharma/gflownet mentioned on GitHubpytorchSyntology: 3 samples, 2 ran (2 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.

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

31 samples harvested: 25 ran (of which 2 constructed an object rather than computing a result; 15 with no instrument failure: 6 honoured, 1 violated, 8 with no contract checked; 10 where Syntology's instrument failed) · 6 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.

6ran · honoured contract
1ran · violated contract
8ran · our draft was wrong
2ran; the contract check could not run
8ran
6unverified

Licence: 18 of the 31 samples are pointer only, meaning Syntology does not serve that copy's text. For 8 of them, this paper's copy is under a permissive licence and cleared, but the stored copy of the same code that Syntology serves is not; the licence label's tooltip gives Syntology's recorded reason. This page shows no code text for any sample; each one links to its file in the repository.

Harvested from 6 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. 4 samples whose link to this paper records no repository are listed under one of this paper's own repositories, because identical code was harvested from it for another paper. 3 samples are listed last, in a group without a repository. “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.

bengioe/gflownet official repository 15 samples, 12 ran (1 instrument-failed) · 3 unverified
array_may_18 mols/ppo.py ran MIT (permissive) · c3caff85ecc30396 · report fixture: signature args:dict/0/? · returned list · level recorded 2026-09-26
branin in 8 papers grid/cond_grid_dag.py ran · honoured contract fingerprinted MIT (permissive) · b8e02a8995424b4e · report fixture: signature x:ndarray/2/float · returned ndarray · level recorded, date not recorded
currin grid/cond_grid_dag.py ran · honoured contract fingerprinted MIT (permissive) · 54dafc6a55c42deb · report fixture: signature x:ndarray/2/float · returned ndarray · level recorded, date not recorded
func_corners grid/toy_grid_dag.py ran · honoured contract fingerprinted MIT (permissive) here; the stored copy is pointer only · bf7643eca4bc2a69 · report fixture: signature x:Tensor/2/float · returned Tensor · level recorded 2026-09-25
func_corners_floor_A grid/toy_grid_dag.py ran · honoured contract fingerprinted MIT (permissive) here; the stored copy is pointer only · 58d90758e97a6d03 · report fixture: signature x:Tensor/2/float · returned Tensor · level recorded 2026-09-25
func_corners_floor_B grid/toy_grid_dag.py ran · honoured contract fingerprinted MIT (permissive) here; the stored copy is pointer only · aa7a567324830eab · report fixture: signature x:Tensor/2/float · returned Tensor · level recorded 2026-09-25
generate_batch in 14 papers grid/toy_grid_dag_al.py ran MIT (permissive) · f08d023272aa048c · report fixture: signature args:Args/0/?, agent:Agent/0/?, dataset:tuple/1/Tensor, env:Env/0/? · returned tuple · level recorded 2026-09-26
get_network_output grid/toy_grid_dag_al.py ran MIT (permissive) · 7ff5d9aaf5641dd3 · report fixture: signature args:Args/0/?, network:callable/0/?, inputs:Tensor/2/float, mean_std:bool/0/bool · returned Tensor · level recorded 2026-09-25
make_mlp in 35 papers grid/cond_grid_dag.py ran · our draft was wrong MIT (permissive) · 2b63a9198ca42b4e · report fixture: signature l:list/1/int, act:callable/0/?, tail:list/1/empty · returned Sequential · level recorded, date not recorded
mol2graph in 5 papers mols/model_block.py ran MIT (permissive) here; the stored copy is pointer only · f763f98f79152e0e · report fixture: signature mol:StandInMol/0/?, mdp:StandInMDP/0/?, floatX:dtype/0/?, bonds:bool/0/bool, nblocks:bool/0/bool · returned Data · level recorded 2026-09-25
mols2batch in 5 papers mols/model_atom.py ran MIT (permissive) here; the stored copy is pointer only · 06477cbf36090384 · report fixture: signature mols:list/1/callable, mdp:MDPConfig/0/? · returned DataBatch · level recorded 2026-09-25
mols2batch in 5 papers mols/model_block.py ran MIT (permissive) here; the stored copy is pointer only · e2f61d62c1a23b4e · report fixture: signature mols:list/1/callable, mdp:MDPStandIn/0/? · returned DataBatch · level recorded 2026-09-25
mols2batch in 5 papers mols/model_fingerprint.py unverified MIT (permissive) here; the stored copy is pointer only · 32208d4977dcb086 · report
train_generative_model mols/gflownet_activelearning.py unverified MIT (permissive) here; the stored copy is pointer only · c29d60fd9bc56a0c · report
train_model_with_proxy mols/gflownet.py unverified MIT (permissive) · df792b9c5dae7a48 · report
SeonghwanSeo/RxnFlow community (archive-listed) 1 sample, 1 ran (0 instrument-failed) · 0 unverified
safe in 5 papers scripts/examples/opt_mogfn.py identical code Syntology harvested from this repository for another paper; this paper's link records no repository, path or licence of its own ran · violated contract licence of this copy not recorded · 1061c8464b631df1 · report fixture: signature f:callable/0/?, x:int/0/int, default:int/0/int · returned int · level recorded, date not recorded
gfnorg/gflownet community (archive-listed) 5 samples, 3 ran (2 instrument-failed) · 2 unverified
GraphAgent mols/model_block.py ran · metamorphic tier: deterministic MIT (permissive) · ece02e732c2290ba · report fixture: returned GraphAgent · level recorded 2026-09-22
array_may_17 mols/mars.py ran · our draft was wrong fingerprinted MIT (permissive) · af9f0f910f9bbc52 · report fixture: signature args:NoneType/0/? · returned list · level recorded, date not recorded
make_mlp in 35 papers grid/cond_grid_dag.py identical code Syntology harvested from this repository for another paper; this paper's link records no repository, path or licence of its own ran · our draft was wrong licence of this copy not recorded · c59313aa188e63c6 · report fixture: signature l:list/1/int, act:callable/0/?, tail:list/1/empty · returned Sequential · level recorded, date not recorded
branin in 8 papers grid/cond_grid_dag.py identical code Syntology harvested from this repository for another paper; this paper's link records no repository, path or licence of its own unverified licence of this copy not recorded · e81128bd6b1cab12 · report
currin grid/cond_grid_dag.py identical code Syntology harvested from this repository for another paper; this paper's link records no repository, path or licence of its own unverified licence of this copy not recorded · 9c972ac9b931bb9e · report
princetonlips/mam community (archive-listed) 3 samples, 3 ran (1 instrument-failed) · 0 unverified
MAM mam/models/mam.py ran no licence file found · pointer only · b78da7a4b43fba13 · report fixture: returned MAM · level recorded 2026-09-22
gen_order mam/models/mam.py ran · honoured contract no licence file found · pointer only · e9f231473aac8e22 · report fixture: signature batch_size:int/0/int, data_dim:int/0/int, device:str/0/str, gen_order:str/0/str · returned Tensor · level recorded 2026-09-22
sample_order mam/models/mam.py ran; the contract check could not run (RAISES) no licence file found · pointer only · afd6888b673d0c17 · report fixture: signature x_batch:Tensor/2/float, gen_order:str/0/str · returned Tensor · level recorded 2026-09-22
recursionpharma/gflownet community (archive-listed) 3 samples, 2 ran (2 instrument-failed) · 1 unverified
cross in 9 papers src/gflownet/algo/trajectory_balance.py ran; the contract check could not run (RAISES) · metamorphic tier: invariant fingerprinted MIT (permissive) · 2a8ed5bf472b7004 · report fixture: signature x:Tensor/1/float · returned Tensor · level recorded, date not recorded
shift_right in 9 papers src/gflownet/algo/trajectory_balance.py ran · our draft was wrong fingerprinted MIT (permissive) · a508d01ed77ced0b · report fixture: signature x:Tensor/2/float, z:int/0/int · returned Tensor · level recorded, date not recorded
subTB in 5 papers src/gflownet/algo/trajectory_balance.py unverified MIT (permissive) · d2e1d2a558b75603 · report
alexhernandezgarcia/gflownet community 1 sample, 1 ran (1 instrument-failed) · 0 unverified
make_opt gflownet/gflownet.py ran · our draft was wrong Apache-2.0 (permissive) · 30d10fc001f71c75 · report fixture: signature params:list/1/Tensor, logZ:list/1/Tensor, config:Config/0/? · returned tuple · level recorded, date not recorded
Repository not located 3 samples, 3 ran (3 instrument-failed) · 0 unverified
3 samples are identical to code Syntology harvested from another paper's repository; this paper's own copy was not located, so no repository is named and no licence is recorded.
binarize_static repository not located ran · our draft was wrong fingerprinted licence of this copy not recorded · 4e16af7e5dbea818 · report fixture: signature x:Tensor/2/float · returned Tensor · level recorded, date not recorded
sample in 167 papers
census: 8 implementations from 9 papers, 7 distinct outputs across 5 buckets, one shared input per bucket
repository not located ran · our draft was wrong licence of this copy not recorded · bf574d554afe6839 · report fixture: signature gfn:callable/0/?, img:Tensor/4/float, p:float/0/float, rand_prob:float/0/float · returned tuple · level recorded, date not recorded
sleep_step repository not located ran · our draft was wrong licence of this copy not recorded · 0461df5888f82edd · report fixture: signature gfn:callable/0/?, decoder:callable/0/?, latent_dict:callable/0/?, batch_size:int/0/int · returned tuple · level recorded, date not recorded

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". 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. 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, listed in 2 groups (siblings lifted in one batch form one group).

generated Assembly for GFlowNet (graph method: GFlowNet), 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, 1 violated, 0 with no contract checked; 0 where Syntology's instrument failed) · 0 unverified. Ran end-to-end against an independent peer (level 3; 6 property tests passed on both).

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

  1. compute_gflownet_reward entry · level 3 · ran · violated contract · ea3452cd70ecb28166e7a205bec389904133afe7bc37a70d32992a6f95f001bc

2 sibling assemblies for GFlowNet (harvested), lifted in the same batch on 2026-09-09: alternative readings kept side by side; none replaces or supersedes another.

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

one of 2 siblings

  1. GraphAgent entry · level 1 · ran · ece02e732c2290ba

harvested Assembly for GFlowNet (graph method: GFlowNet). 3 bricks; individually on a fixture: 3 ran (of which 1 constructed an object rather than computing a result; 2 with no instrument failure: 1 honoured, 0 violated, 1 with no contract checked; 1 where Syntology's instrument failed) · 0 unverified. The assembly has not been run end-to-end.

one of 2 siblings

  1. MAM entry · level 1 · ran · b78da7a4b43fba13
  2. gen_order helper · level 1 · ran · honoured contract · e9f231473aac8e22
  3. sample_order helper · level 1 · ran; the contract check could not run (RAISES) · afd6888b673d0c17

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="2106.04399") returns these assemblies with each brick's role and wiring (how to connect).

Results from the paper archive 2025-07-28

No leaderboard rows for this paper in the archive.

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