pith. machine review for the scientific record. sign in

arxiv: 2604.05075 · v1 · submitted 2026-04-06 · 💻 cs.AI · cs.CL

Recognition: 2 theorem links

· Lean Theorem

MMORF: A Multi-agent Framework for Designing Multi-objective Retrosynthesis Planning Systems

Frazier N. Baker , Trieu Nguyen , Reza Averly , Botao Yu , Daniel Adu-Ampratwum , Huan Sun , Xia Ning
Authors on Pith no claims yet

Pith reviewed 2026-05-10 19:43 UTC · model grok-4.3

classification 💻 cs.AI cs.CL
keywords multi-agent systemsretrosynthesis planningmulti-objective optimizationlanguage modelschemical synthesisAI for chemistryagent interactions
0
0 comments X

The pith

A modular multi-agent framework lets language models balance safety, cost and quality in retrosynthesis planning.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper introduces MMORF as a framework for assembling specialized language-model agents into systems that handle multiple conflicting objectives during retrosynthesis planning. It supplies modular components that can be combined in different ways, allowing systematic tests of alternative designs. On a benchmark of 218 tasks, one resulting system achieves strong safety and cost results on soft-constraint problems and often produces routes that dominate baselines in trade-offs, while a second system reaches 48.6 percent success on hard-constraint problems and exceeds prior methods. A reader would care because retrosynthesis planning underpins chemical synthesis and drug design, where quality, safety, and cost must be weighed together. The framework's value lies in turning agent interactions into a configurable tool for this balance.

Core claim

MMORF supplies modular agentic components that can be flexibly assembled into multi-agent systems for multi-objective retrosynthesis planning; when instantiated as MASIL and RFAS, these systems deliver improved safety and cost metrics on soft-constraint tasks and a 48.6 percent success rate on hard-constraint tasks, outperforming existing baselines on the 218-task benchmark.

What carries the argument

MMORF, a framework of modular agentic components that are combined and configured into different multi-agent systems to incorporate multiple objectives into retrosynthesis routes.

If this is right

  • Different multi-agent system designs for retrosynthesis can be evaluated and compared in a principled way through MMORF's modular structure.
  • MASIL frequently produces routes that Pareto-dominate baseline routes on soft-constraint tasks for safety and cost.
  • RFAS reaches a 48.6 percent success rate on hard-constraint tasks and exceeds state-of-the-art baselines.
  • The framework supports further exploration of language-model agent architectures for balancing quality, safety, and cost in chemical planning.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • The same modular agent approach could be tested on multi-objective planning problems outside chemistry, such as supply-chain or energy-resource allocation.
  • Adding mechanisms for agents to incorporate live experimental feedback might reduce route inconsistencies that the current framework leaves unmeasured.
  • Scaling the benchmark beyond 218 tasks would reveal whether the observed performance gaps hold under broader chemical diversity.

Load-bearing premise

Interactions among specialized language-model agents can reliably balance multiple conflicting objectives without introducing unmeasured biases or inconsistencies in the generated routes.

What would settle it

A re-evaluation of the 218-task benchmark that finds the reported safety and cost metrics for MASIL or the 48.6 percent success rate for RFAS to be overstated or accompanied by overlooked safety violations would disprove the performance claims.

Figures

Figures reproduced from arXiv: 2604.05075 by Botao Yu, Daniel Adu-Ampratwum, Frazier N. Baker, Huan Sun, Reza Averly, Trieu Nguyen, Xia Ning.

Figure 1
Figure 1. Figure 1: Overview of a, MMORF, and two MAS built using it: b, MASIL, and c, RFAS. HCMO-retro with a single constraint. However, such formulations do not capture more realis￾tic settings, such as HCMO-retro with multiple constraints, or soft-constraint multi-objective retrosynthesis planning (SCMO-retro)—where many objectives must be dynamically balanced throughout planning. MAS are well-suited to these challenging … view at source ↗
Figure 2
Figure 2. Figure 2: Synthesis of thiothixene generated by MASIL for a [PITH_FULL_IMAGE:figures/full_fig_p009_2.png] view at source ↗
read the original abstract

Multi-objective retrosynthesis planning is a critical chemistry task requiring dynamic balancing of quality, safety, and cost objectives. Language model-based multi-agent systems (MAS) offer a promising approach for this task: leveraging interactions of specialized agents to incorporate multiple objectives into retrosynthesis planning. We present MMORF, a framework for constructing MAS for multi-objective retrosynthesis planning. MMORF features modular agentic components, which can be flexibly combined and configured into different systems, enabling principled evaluation and comparison of different system designs. Using MMORF, we construct two representative MAS: MASIL and RFAS. On a newly curated benchmark consisting of 218 multi-objective retrosynthesis planning tasks, MASIL achieves strong safety and cost metrics on soft-constraint tasks, frequently Pareto-dominating baseline routes, while RFAS achieves a 48.6% success rate on hard-constraint tasks, outperforming state-of-the-art baselines. Together, these results show the effectiveness of MMORF as a foundational framework for exploring MAS for multi-objective retrosynthesis planning. Code and data are available at https://anonymous.4open.science/r/MMORF/.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

2 major / 1 minor

Summary. The paper introduces MMORF, a modular framework for constructing multi-agent systems (MAS) tailored to multi-objective retrosynthesis planning. It uses MMORF to instantiate two systems (MASIL for soft constraints and RFAS for hard constraints) and evaluates them on a newly curated benchmark of 218 tasks. The central claims are that MASIL frequently Pareto-dominates baseline routes on safety and cost metrics for soft-constraint tasks, while RFAS achieves a 48.6% success rate on hard-constraint tasks and outperforms state-of-the-art baselines. The work positions MMORF as a foundational tool for exploring MAS designs in this domain, with code and data released.

Significance. If the results hold, MMORF offers a systematic way to design and compare MAS for balancing conflicting objectives (quality, safety, cost) in retrosynthesis, an important application of AI in chemistry. The open code and data release is a clear strength that enables reproducibility and extension. The empirical gains on a dedicated benchmark could advance multi-objective planning if they demonstrably arise from principled agent coordination rather than LLM-specific artifacts.

major comments (2)
  1. [Abstract and experimental evaluation section] The performance claims (MASIL Pareto dominance on soft constraints; RFAS 48.6% success on hard constraints) rest on the premise that specialized LLM agents dynamically reconcile conflicting objectives via interaction. The manuscript provides no validation of chemical validity of generated routes (e.g., template correctness or intermediate feasibility) or consistency of outputs across repeated runs of the same task. Aggregate metrics alone cannot distinguish principled multi-objective balancing from training-data biases or inconsistencies, which is load-bearing for the framework's claimed value.
  2. [Benchmark and evaluation section] The benchmark of 218 tasks is described as newly curated, yet the manuscript lacks detail on task selection criteria, data sources, definition of soft- vs. hard-constraint splits, or how baselines were implemented and tuned. Without these, it is difficult to assess whether the reported improvements are robust or generalizable.
minor comments (1)
  1. [Abstract] The anonymous code link should be replaced with a permanent repository (e.g., GitHub or Zenodo) in the camera-ready version to support the reproducibility claim.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for their constructive and detailed comments, which highlight important aspects for improving the clarity, rigor, and reproducibility of our work on MMORF. We address each major comment point by point below, outlining specific revisions we will make to the manuscript.

read point-by-point responses

  1. Referee: [Abstract and experimental evaluation section] The performance claims (MASIL Pareto dominance on soft constraints; RFAS 48.6% success on hard constraints) rest on the premise that specialized LLM agents dynamically reconcile conflicting objectives via interaction. The manuscript provides no validation of chemical validity of generated routes (e.g., template correctness or intermediate feasibility) or consistency of outputs across repeated runs of the same task. Aggregate metrics alone cannot distinguish principled multi-objective balancing from training-data biases or inconsistencies, which is load-bearing for the framework's claimed value.

    Authors: We agree that explicit validation of chemical validity and output consistency is essential to substantiate claims about the benefits of agent interactions in reconciling objectives. The original manuscript follows common practice in retrosynthesis literature by reporting aggregate metrics (success rate, safety, cost) on top of established base models assumed to produce valid outputs. However, this does not fully address potential biases or variability. In the revised version, we will add a dedicated subsection under experimental evaluation that includes: (1) manual chemical validity checks (template correctness and intermediate feasibility) on a random sample of 20 routes by domain experts, and (2) consistency analysis by rerunning a subset of 10 tasks across 5 independent runs with varied seeds, reporting metric variance. These additions will help isolate the contribution of the multi-agent coordination. We will also explicitly discuss remaining limitations regarding LLM-specific artifacts. revision: yes

  2. Referee: [Benchmark and evaluation section] The benchmark of 218 tasks is described as newly curated, yet the manuscript lacks detail on task selection criteria, data sources, definition of soft- vs. hard-constraint splits, or how baselines were implemented and tuned. Without these, it is difficult to assess whether the reported improvements are robust or generalizable.

    Authors: We concur that greater transparency on benchmark construction is required to support claims of robustness and generalizability. The 218 tasks were derived from public retrosynthesis datasets (primarily USPTO-derived sources), with selection criteria emphasizing molecules that present clear multi-objective trade-offs. Soft-constraint tasks permit balanced satisfaction of objectives, while hard-constraint tasks enforce strict satisfaction of all constraints. Baselines were reimplemented using the identical LLM backbone as MASIL/RFAS, with hyperparameters tuned via grid search on a validation split. In the revision, we will expand the benchmark section with a new subsection detailing task selection criteria, exact data sources, precise definitions of the soft/hard splits, and baseline implementation/tuning procedures (including pseudocode). The already-released code and data repository will be updated with the corresponding construction scripts and documentation to enable full reproducibility. revision: yes

Circularity Check

0 steps flagged

No circularity: empirical framework evaluation on external benchmark

full rationale

The paper defines MMORF as a modular framework for composing multi-agent systems, instantiates MASIL and RFAS, and reports measured success rates (48.6% for RFAS on hard constraints) and Pareto dominance on a newly curated 218-task benchmark. These outcomes are obtained by direct comparison against external baselines rather than by any internal equation, fitted parameter, or self-citation that reduces the result to its own inputs. No self-definitional loops, renamed predictions, or load-bearing uniqueness theorems appear in the derivation chain.

Axiom & Free-Parameter Ledger

0 free parameters · 0 axioms · 0 invented entities

Abstract provides no explicit free parameters, axioms, or invented entities; the framework is described at the level of modular agentic components without detailing internal fitting procedures or new postulated mechanisms.

pith-pipeline@v0.9.0 · 5522 in / 1085 out tokens · 48072 ms · 2026-05-10T19:43:19.712824+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Lean theorems connected to this paper

Citations machine-checked in the Pith Canon. Every link opens the source theorem in the public Lean library.

What do these tags mean?
matches
The paper's claim is directly supported by a theorem in the formal canon.
supports
The theorem supports part of the paper's argument, but the paper may add assumptions or extra steps.
extends
The paper goes beyond the formal theorem; the theorem is a base layer rather than the whole result.
uses
The paper appears to rely on the theorem as machinery.
contradicts
The paper's claim conflicts with a theorem or certificate in the canon.
unclear
Pith found a possible connection, but the passage is too broad, indirect, or ambiguous to say the theorem truly supports the claim.

Reference graph

Works this paper leans on

34 extracted references · 23 canonical work pages · 3 internal anchors

  1. [1]

    write newline

    " write newline "" before.all 'output.state := FUNCTION n.dashify 't := "" t empty not t #1 #1 substring "-" = t #1 #2 substring "--" = not "--" * t #2 global.max substring 't := t #1 #1 substring "-" = "-" * t #2 global.max substring 't := while if t #1 #1 substring * t #2 global.max substring 't := if while FUNCTION format.date year duplicate empty "emp...

  2. [2]

    @esa (Ref

    \@ifxundefined[1] #1\@undefined \@firstoftwo \@secondoftwo \@ifnum[1] #1 \@firstoftwo \@secondoftwo \@ifx[1] #1 \@firstoftwo \@secondoftwo [2] @ #1 \@temptokena #2 #1 @ \@temptokena \@ifclassloaded agu2001 natbib The agu2001 class already includes natbib coding, so you should not add it explicitly Type <Return> for now, but then later remove the command n...

  3. [3]

    \@lbibitem[] @bibitem@first@sw\@secondoftwo \@lbibitem[#1]#2 \@extra@b@citeb \@ifundefined br@#2\@extra@b@citeb \@namedef br@#2 \@nameuse br@#2\@extra@b@citeb \@ifundefined b@#2\@extra@b@citeb @num @parse #2 @tmp #1 NAT@b@open@#2 NAT@b@shut@#2 \@ifnum @merge>\@ne @bibitem@first@sw \@firstoftwo \@ifundefined NAT@b*@#2 \@firstoftwo @num @NAT@ctr \@secondoft...

  4. [4]

    @open @close @open @close and [1] URL: #1 \@ifundefined chapter * \@mkboth \@ifxundefined @sectionbib * \@mkboth * \@mkboth\@gobbletwo \@ifclassloaded amsart * \@ifclassloaded amsbook * \@ifxundefined @heading @heading NAT@ctr thebibliography [1] @ \@biblabel @NAT@ctr \@bibsetup #1 @NAT@ctr @ @openbib .11em \@plus.33em \@minus.07em 4000 4000 `\.\@m @bibit...

  5. [5]

    Introducing Claude 4.5 Sonnet , 2025

    Anthropic. Introducing Claude 4.5 Sonnet , 2025. URL https://www.anthropic.com/news/claude-sonnet-4-5

    2025

  6. [6]

    Tango*: Constrained synthesis planning using chemically informed value functions, 2024

    Daniel Armstrong, Zlatko Joncev, Jeff Guo, and Philippe Schwaller. Tango*: Constrained synthesis planning using chemically informed value functions, 2024. URL https://arxiv.org/abs/2412.03424

    work page arXiv 2024

  7. [7]

    Liddia: Language-based intelli- gent drug discovery agent,

    Reza Averly, Frazier N. Baker, and Xia Ning. LIDDIA : Language -based Intelligent Drug Discovery Agent , February 2025. URL http://arxiv.org/abs/2502.13959. arXiv:2502.13959 [cs]

    work page arXiv 2025

  8. [8]

    Baker, Ziqi Chen, Daniel Adu-Ampratwum, and Xia Ning

    Frazier N. Baker, Ziqi Chen, Daniel Adu-Ampratwum, and Xia Ning. Rlsync: Offline-online reinforcement learning for synthon completion, 2024. URL https://arxiv.org/abs/2309.02671

    work page arXiv 2024

  9. [9]

    Baker, Daniel Adu-Ampratwum, Reza Averly, Botao Yu, Huan Sun, and Xia Ning

    Frazier N. Baker, Daniel Adu-Ampratwum, Reza Averly, Botao Yu, Huan Sun, and Xia Ning. Larc: Towards human-level constrained retrosynthesis planning through an agentic framework, 2025. URL https://arxiv.org/abs/2508.11860

    work page arXiv 2025

  10. [10]

    Bran, Theo A

    Andres M. Bran, Theo A. Neukomm, Daniel P. Armstrong, Zlatko Jončev, and Philippe Schwaller. Chemical reasoning in LLMs unlocks steerable synthesis planning and reaction mechanism elucidation, March 2025. URL http://arxiv.org/abs/2503.08537. arXiv:2503.08537 [cs]

    work page arXiv 2025

  11. [11]

    Route Design and Selection

    Mike Butters. Route Design and Selection . In Pharmaceutical Process Development - Current Chemical and Engineering Challenges , pp.\ 90--116. Royal Society of Chemistry (RSC), 2011. ISBN 978-1-84973-146-1. URL https://app.knovel.com/hotlink/pdf/id:kt00A9X19B/pharmaceutical-process/throughput

    2011

  12. [12]

    Retro*: Learning retrosynthetic planning with neural guided a* search

    Binghong Chen, Chengtao Li, Hanjun Dai, and Le Song. Retro*: Learning retrosynthetic planning with neural guided a* search. In The 37th International Conference on Machine Learning (ICML 2020), 2020

    2020

  13. [13]

    Retrogfn: Diverse and feasible retrosynthesis using gflownets, 2024

    Piotr Gaiński, Michał Koziarski, Krzysztof Maziarz, Marwin Segler, Jacek Tabor, and Marek Śmieja. Retrogfn: Diverse and feasible retrosynthesis using gflownets, 2025. URL https://arxiv.org/abs/2406.18739

    work page arXiv 2025

  14. [14]

    doi:10.1186/s13321-020-00472-1 , pages =

    Samuel Genheden, Amol Thakkar, Veronika Chadimová, Jean-Louis Reymond, Ola Engkvist, and Esben Bjerrum. AiZynthFinder : a fast, robust and flexible open-source software for retrosynthetic planning. Journal of Cheminformatics, 12 0 (1): 0 70, November 2020. ISSN 1758-2946. doi:10.1186/s13321-020-00472-1. URL https://doi.org/10.1186/s13321-020-00472-1

    work page doi:10.1186/s13321-020-00472-1 2020

  15. [15]

    and Weber, Joanne D

    Gibson, Jack R. and Weber, Joanne D. Handbook of Selected Properties of Air - and Water - Reactive Materials . Technical Report 144, Defense Technical Information Center, Crane, Indiana, March 1969. URL http://archive.org/details/DTIC_AD0688422

    1969

  16. [16]

    Available: http://dx.doi.org/10.1038/s41586-025-09422-z

    Daya Guo, Dejian Yang, Haowei Zhang, Junxiao Song, Peiyi Wang, Qihao Zhu, et al. Deepseek-r1 incentivizes reasoning in llms through reinforcement learning. Nature, 645 0 (8081): 0 633–638, September 2025. ISSN 1476-4687. doi:10.1038/s41586-025-09422-z. URL http://dx.doi.org/10.1038/s41586-025-09422-z

    work page doi:10.1038/s41586-025-09422-z 2025

  17. [17]

    Coley, and Ying Wei

    Yinjie Jiang, Yemin Yu, Ming Kong, Yu Mei, Luotian Yuan, Zhengxing Huang, Kun Kuang, Zhihua Wang, Huaxiu Yao, James Zou, Connor W. Coley, and Ying Wei. Artificial Intelligence for Retrosynthesis Prediction . Engineering, 25: 0 32--50, June 2023. ISSN 2095-8099. doi:10.1016/j.eng.2022.04.021. URL https://www.sciencedirect.com/science/article/pii/S2095809922005665

    work page doi:10.1016/j.eng.2022.04.021 2023

  18. [18]

    Self-improved retrosynthetic planning, 2021

    Junsu Kim, Sungsoo Ahn, Hankook Lee, and Jinwoo Shin. Self-improved retrosynthetic planning, 2021. URL https://arxiv.org/abs/2106.04880

    work page arXiv 2021

  19. [19]

    Thiessen, Tiejun Cheng, Bo Yu, and Evan E

    Sunghwan Kim, Paul A. Thiessen, Tiejun Cheng, Bo Yu, and Evan E. Bolton. An update on PUG - REST : RESTful interface for programmatic access to PubChem . Nucleic Acids Research, 46 0 (W1): 0 W563--W570, July 2018. ISSN 1362-4962. doi:10.1093/nar/gky294

    work page doi:10.1093/nar/gky294 2018

  20. [20]

    and Clevert, Djork-Arné and Preuss, Mike and Genheden, Samuel , year =

    Helen Lai, Christos Kannas, Alan Kai Hassen, Emma Granqvist, Annie M. Westerlund, Djork-Arné Clevert, Mike Preuss, and Samuel Genheden. Multi-objective synthesis planning by means of Monte Carlo Tree search. Artificial Intelligence in the Life Sciences, 7: 0 100130, June 2025. ISSN 2667-3185. doi:10.1016/j.ailsci.2025.100130. URL https://www.sciencedirect...

    work page doi:10.1016/j.ailsci.2025.100130 2025

  21. [21]

    Chemical reactions from US patents (1976- Sep2016 ), 2017

    Daniel Lowe. Chemical reactions from US patents (1976- Sep2016 ), 2017. URL https://figshare.com/articles/dataset/Chemical_reactions_from_US_patents_1976-Sep2016_/5104873/1. Artwork Size: 1494665893 Bytes Pages: 1494665893 Bytes

    1976

  22. [22]

    Mistral NeMo , July 2024

    Mistral AI . Mistral NeMo , July 2024. URL https://mistral.ai/news/mistral-nemo

    2024

  23. [23]

    H. L. Morgan. The Generation of a Unique Machine Description for Chemical Structures - A Technique Developed at Chemical Abstracts Service . Journal of Chemical Documentation, 5 0 (2): 0 107--113, May 1965. ISSN 0021-9576. doi:10.1021/c160017a018. URL https://doi.org/10.1021/c160017a018

    work page doi:10.1021/c160017a018 1965

  24. [24]

    RDKit : Open -source cheminformatics

    RDKit. RDKit : Open -source cheminformatics. URL https://www.rdkit.org

  25. [25]

    Robinson, Alpha Lee, Frank Von Delft, and Charlotte M

    Ruben Sanchez-Garcia, Dávid Havasi, Gergely Takács, Matthew C. Robinson, Alpha Lee, Frank Von Delft, and Charlotte M. Deane. CoPriNet : graph neural networks provide accurate and rapid compound price prediction for molecule prioritisation. Digital Discovery, 2 0 (1): 0 103--111, 2023. ISSN 2635-098X. doi:10.1039/D2DD00071G. URL https://xlink.rsc.org/?DOI=...

    work page doi:10.1039/d2dd00071g 2023

  26. [26]

    Marwin H. S. Segler, Mike Preuss, and Mark P. Waller. Planning chemical syntheses with deep neural networks and symbolic AI . Nature, 555 0 (7698): 0 604--610, March 2018. ISSN 1476-4687. doi:10.1038/nature25978. URL https://www.nature.com/articles/nature25978. Publisher: Nature Publishing Group

    work page doi:10.1038/nature25978 2018

  27. [27]

    OpenAI GPT-5 System Card

    Aaditya Singh, Adam Fry, Adam Perelman, Adam Tart, Adi Ganesh, Ahmed El-Kishky, et al. Openai gpt-5 system card, 2025. URL https://arxiv.org/abs/2601.03267

    work page internal anchor Pith review Pith/arXiv arXiv 2025

  28. [28]

    Tobias Sutter, Bart P. G. Van Parys, and Daniel Kuhn. A pareto dominance principle for data-driven optimization, 2023. URL https://arxiv.org/abs/2010.06606

    work page arXiv 2023

  29. [29]

    ADMET - AI : a machine learning ADMET platform for evaluation of large-scale chemical libraries

    Kyle Swanson, Parker Walther, Jeremy Leitz, Souhrid Mukherjee, Joseph C Wu, Rabindra V Shivnaraine, and James Zou. ADMET - AI : a machine learning ADMET platform for evaluation of large-scale chemical libraries. Bioinformatics, 40 0 (7): 0 btae416, July 2024. ISSN 1367-4811. doi:10.1093/bioinformatics/btae416. URL https://doi.org/10.1093/bioinformatics/btae416

    work page doi:10.1093/bioinformatics/btae416 2024

  30. [30]

    Kimi Team, Tongtong Bai, Yifan Bai, Yiping Bao, S. H. Cai, Yuan Cao, et al. Kimi k2.5: Visual agentic intelligence, 2026. URL https://arxiv.org/abs/2602.02276

    work page internal anchor Pith review Pith/arXiv arXiv 2026

  31. [31]

    and Jončev, Zlatko and Schwaller, Philippe , year =

    Nguyen Xuan-Vu, Daniel Armstrong, Milena Wehrbach, Andres M. Bran, Zlatko Jončev, and Philippe Schwaller. Synthelite: Chemist -aligned and feasibility-aware synthesis planning with LLMs , December 2025. URL http://arxiv.org/abs/2512.16424. arXiv:2512.16424 [cs]

    work page arXiv 2025

  32. [32]

    Qwen3 Technical Report

    An Yang, Anfeng Li, Baosong Yang, Beichen Zhang, Binyuan Hui, Bo Zheng, et al. Qwen3 technical report, 2025. URL https://arxiv.org/abs/2505.09388

    work page internal anchor Pith review Pith/arXiv arXiv 2025

  33. [33]

    Kevin Yu, Jihye Roh, Ziang Li, Wenhao Gao, Runzhong Wang, and Connor W. Coley. Double-ended synthesis planning with goal-constrained bidirectional search, 2024. URL https://arxiv.org/abs/2407.06334

    work page arXiv 2024

  34. [34]

    Communications Chemistry , author =

    Dengwei Zhao, Shikui Tu, and Lei Xu. Efficient retrosynthetic planning with MCTS exploration enhanced A * search. Commun Chem, 7 0 (1): 0 1--12, March 2024. ISSN 2399-3669. doi:10.1038/s42004-024-01133-2. URL https://www.nature.com/articles/s42004-024-01133-2. Publisher: Nature Publishing Group

    work page doi:10.1038/s42004-024-01133-2 2024