Pith. sign in

REVIEW 4 major objections 4 minor 1 cited by

Uni-Mol3: A Multi-Molecular Foundation Model for Advancing Organic Reaction Modeling

T0 review · 4 major / 4 minor · reviewed 2026-08-06 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

Pith's one-line read Uni-Mol3 claims a 3D-aware tokenizer plus two-stage pre-training lets one model handle entire reacting systems and beat prior methods on reaction tasks.

desk verdict A plausible and significant extension of the Uni-Mol line, but the empirical core is unverifiable from the text I received. read the letter →

arxiv 2508.00920 v2 pith:FP46I7W3 submitted 2025-07-30 physics.chem-ph cs.LG

Lirong Wu , Junjie Wang , Zhifeng Gao , Xiaohong Ji , Rong Zhu , Xinyu Li , Linfeng Zhang , Guolin Ke
show 1 more author
Weinan E
This is my paper · ORCID
classification physics.chem-phcs.LG
keywords Uni-Mol3organicreactionmodelingmulti-molecularfoundationmodel3D-awaretokenizationmolecularpre-trainingrepresentationlearningtransformer
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

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

The reading

The paper is trying to establish that a foundation model built for whole reacting systems, not just isolated molecules, can capture organic reaction behavior better than existing single-molecule models. It introduces Uni-Mol3, which discretizes 3D molecular structure into tokens via a Mol-Tokenizer, pre-trains first on molecules and then on reactions, and fine-tunes with prompt-aware adapters. Across 10 datasets and 4 downstream reaction tasks, the authors report consistent gains over existing methods, including the single-molecular predecessor Uni-Mol2. If true, this means reaction modeling benefits from learning molecular grammar and reaction principles jointly in one model rather than treating each molecule separately.

What carries the argument

The load-bearing object is the Mol-Tokenizer, a multi-scale molecular tokenizer that encodes 3D structures of molecules and other features into discrete tokens, producing a vocabulary of 3D-aware molecular tokens. This tokenization is what makes multi-molecular systems tractable as sequences for the transformer backbone, and the two-stage pre-training (molecular pre-training for molecular grammars, reaction pre-training for reaction principles) is what transfers the representation to reaction tasks, with prompt-aware downstream fine-tuning adapting the model to specific tasks.

What would settle it

Find a reaction dataset where the outcome is governed by a fine 3D detail, such as stereoselectivity set by a chiral center's exact spatial arrangement or a specific conformational preference, and show that Uni-Mol3's tokenized representation maps two molecules differing only in that detail to the same token sequence; if the model then predicts identical outcomes, the central claim fails. Alternatively, an ablation that removes the reaction pre-training stage while retaining nearly all downstream performance would undercut the progressive-learning explanation.

Watch

Extended reading notes

Core claim

On its own terms, the paper's central discovery is that discretizing 3D molecular geometry into a token vocabulary creates a 3D-aware molecular language that, when pre-trained in two stages (molecules first, reactions second), transfers to a range of organic reaction tasks. The claim is that this hierarchical pipeline lets the model handle multi-molecular systems directly and outperforms existing single-molecular representation models on the evaluated benchmarks. The authors assert this validates a progressive learning paradigm from single-molecular to multi-molecular systems.

Load-bearing premise

The design assumes that quantizing molecular 3D structure into a finite token vocabulary keeps the stereochemical, geometric, and electronic details that decide how a reaction actually proceeds.

Editorial extensions

If this is right

  • Reaction prediction, retrosynthesis, and other organic reaction tasks can be served by a single pre-trained multi-molecular model instead of task-specific single-molecule encoders.
  • The two-stage pre-training order, molecules first then reactions, acts as an effective curriculum for learning reaction principles.
  • Discretizing 3D geometry into tokens does not destroy the information needed for downstream reaction modeling, given the reported gains.
  • Multi-task prediction with strong generalizability is achievable without retraining the backbone for each task.
  • The approach establishes an alternative paradigm for multi-molecular computational chemistry beyond single-molecular representation learning.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • If the tokenizer truly preserves stereo and electronic detail, the same discrete-token recipe could extend to reaction condition prediction, catalyst design, or selectivity modeling, where 3D detail is decisive.
  • The claim would be sharpened by an ablation showing the reaction pre-training stage contributes beyond molecular pre-training alone; the paper's framing implies this but the abstract does not quantify it.
  • A direct test would compare Uni-Mol3 against a version using continuous 3D features rather than tokenization on stereochemistry-sensitive tasks, isolating what discretization buys.
  • The paradigm hints that language-model-style scaling laws may apply to reaction data, with more pre-training reactions yielding better mechanistic understanding downstream.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

4 major / 4 minor

Summary. The paper introduces Uni-Mol3, a multi-molecular foundation model for organic reaction modeling. The proposed architecture combines a multi-scale Mol-Tokenizer that encodes 3D structures into discrete tokens, two pre-training stages (molecular and reaction), and prompt-aware downstream fine-tuning. The abstract claims that Uni-Mol3 outperforms existing methods across 10 datasets spanning 4 downstream tasks, and the authors position the framework as a bridge between molecular representation learning and reaction mechanism understanding. The supplied full text is severely corrupted, so the only reliably readable content is the abstract; the body, tables, and references are unusable as provided.

Significance. If the stated claims were fully supported, Uni-Mol3 would be a notable step toward extending molecular foundation models from single-molecule representations to multi-molecular reaction modeling. The idea of a discrete, 3D-aware molecular tokenizer followed by two-stage pre-training is a plausible and potentially useful design direction, and the use of external reaction benchmarks would give the central empirical claim independent grounding. However, no quantitative results, baseline identifiers, dataset names, error bars, or ablations are visible in the readable portion of the manuscript, and the full text cannot be evaluated. The significance of the contribution is therefore currently unassessable, and the paper provides no reproducible artifacts or machine-checked derivations to support its claims.

major comments (4)
  1. [Abstract] The central claim that Uni-Mol3 "outperforms existing methods" across 10 datasets and 4 downstream tasks is stated without a single quantitative result, baseline name, dataset identifier, metric, or uncertainty estimate. Since the experimental tables in the supplied full text are unreadable, this sentence is the only evidence available for the paper's main conclusion, and it is insufficient to support that conclusion.
  2. [Full text] The supplied manuscript body is irrecoverably corrupted, with repeated mojibake and an embedded header from an unrelated paper (arXiv:2508.00918v1 [astro-ph.IM]). This makes it impossible to review the methodology, experimental setup, results, or ablations, and the empirical claim in the abstract is therefore not checkable from the submitted artifact.
  3. [Abstract (Mol-Tokenizer description)] The design premise that discretizing 3D molecular structures into a finite token vocabulary preserves the stereochemical and electronic information that determines reaction outcomes is asserted but not demonstrated. The manuscript provides no vocabulary size, quantization-error analysis, conformer-quality validation, or reconstruction experiment; without such support, the mechanism attributed to the Mol-Tokenizer remains an unverified assumption.
  4. [Abstract (two-stage pre-training)] The claimed benefit of molecular pre-training followed by reaction pre-training is presented as the key contribution, but no ablation isolates the contribution of either stage. Because the paper's causal narrative rests on this progressive learning paradigm, an ablation removing each stage, or reversing their order, is load-bearing evidence that the provided text does not contain.
minor comments (4)
  1. [Abstract] The phrase "multi-task prediction with strong generalizability" is a qualitative claim; the paper should specify the tasks, the generalization protocol, and the quantities that support this assertion.
  2. [Abstract] The abstract mentions "10 datasets spanning 4 downstream tasks" but does not name the datasets or tasks; these should be listed, along with evaluation metrics and data splits, either in the abstract or in a clearly readable summary table.
  3. [Full text] The final pages contain repeated and garbled content that appears to include the reference list, so the related-work positioning and prior-art attribution cannot be checked in the submitted version.
  4. [Full text] No code or data availability statement is visible; for a foundation-model paper, releasing the tokenizer, pre-training checkpoints, and downstream fine-tuning code would substantially aid reproducibility and verification.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity identified from the available abstract; the full text is corrupted and cannot provide a derivation chain to inspect.

full rationale

The only readable portion of the manuscript is the abstract, which reports that Uni-Mol3 outperforms existing methods on 10 datasets spanning 4 downstream tasks. These benchmarks are external reaction datasets, so the central performance claim has independent grounding and is not defined in terms of the model's own outputs. The abstract does name Uni-Mol2 as prior single-molecular work, and Uni-Mol3 is evidently a successor from the same research line, but no load-bearing argument is visibly reduced to a self-citation: the paper does not invoke a uniqueness theorem, does not call a fitted parameter a prediction, and does not define its tokenizer or pretraining objectives in terms of the benchmark outcomes within the supplied text. The supplied full text is UTF-8-corrupted and even contains an unrelated arXiv header, so the detailed equations, benchmark tables, and ablation comparisons cannot be inspected. That corruption makes the empirical claims unverifiable from this artifact, but unverifiability is not circularity. No specific equation, definition, or citation chain can be exhibited that reduces a claimed result to its own inputs. Accordingly, the appropriate circularity finding is a clean non-finding at score 0.

Assumptions & free parameters 3 free parameters · 4 assumptions · 2 invented entities

Because only the abstract was readable, the ledger is necessarily incomplete. The listed free parameters are the ones implied by the abstract's architecture description (tokenization scale, pre-training balance, model size); all are fitted or hand-chosen, none derived. The axioms are domain assumptions about conformer quality, tokenizer information retention, and transfer; the central claim depends on all three. Mol-Tokenizer and the reaction pre-training stage are the only genuinely new constructs, and neither has independent evidence outside the paper's reported benchmarks.

free parameters (3)
  • Mol-Tokenizer vocabulary size and scale count = not reported in abstract
    The number of discrete tokens and multi-scale levels determines how much 3D detail survives encoding; a hand-chosen design constant that the performance claim depends on.
  • Pre-training loss weights (molecular vs. reaction stage) = not reported in abstract
    The balance between the two pre-training objectives is tuned, not derived, and is invisible at abstract level.
  • Architecture hyperparameters (layers, hidden dim, attention heads) = not reported in abstract
    Standard foundation model hyperparameters presumably selected on validation sets of the same benchmark family.
assumptions (4)
  • domain assumption Conformer structures fed to the tokenizer are accurate enough for reaction prediction.
    Reaction corpora do not provide experimental 3D structures; generated conformers carry geometric noise that the model must tolerate.
  • domain assumption Discrete tokenization preserves reaction-critical chemical information.
    The entire architecture rests on the premise that irreversible tokenization of 3D coordinates loses no information that determines reaction outcomes.
  • domain assumption Single-molecule pre-training transfers positively to multi-molecular reaction tasks.
    The progressive pre-training paradigm assumes the molecular stage helps the reaction stage; no no-pretraining ablation is visible in the abstract.
  • standard math Standard transformer and attention machinery behaves as expected for tokenized molecular sequences.
    The model builds on standard deep learning building blocks without modification.
invented entities (2)
  • Mol-Tokenizer (multi-scale molecular tokenizer)
    purpose: Encodes 3D molecular structures and features into discrete tokens, forming a 3D-aware molecular language for the backbone model.
    A new model component whose benefit is observable only through the paper's own benchmark suite; no external falsifiable handle.
  • Reaction pre-training stage
    purpose: Captures reaction principles by training on multi-molecular systems before prompt-aware fine-tuning on downstream tasks.
    An internal training procedure; its contribution is only evidenced by the reported benchmarks, not by any external test.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Uni-Mol3: A Multi-Molecular Foundation Model for Advancing Organic Reaction Modeling." pith.science (2026). https://pith.science/paper/FP46I7W3

@misc{pith2026250800920,
  author       = {Pith},
  title        = {Pith review of: Uni-Mol3: A Multi-Molecular Foundation Model for Advancing Organic Reaction Modeling},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/FP46I7W3}},
  note         = {Machine review of arXiv:2508.00920}
}
read the original abstract

Organic reaction, the foundation of modern chemical industry, is crucial for new material development and drug discovery. However, deciphering reaction mechanisms and modeling multi-molecular relationships remain formidable challenges due to the complexity of molecular dynamics. While several state-of-the-art models like Uni-Mol2 have revolutionized single-molecular representation learning, their extension to multi-molecular systems, where chemical reactions inherently occur, has been underexplored. This paper introduces Uni-Mol3, a novel deep learning framework that employs a hierarchical pipeline for multi-molecular reaction modeling. At its core, Uni-Mol3 adopts a multi-scale molecular tokenizer (Mol-Tokenizer) that encodes 3D structures of molecules and other features into discrete tokens, creating a 3D-aware molecular language. The framework innovatively combines two pre-training stages: molecular pre-training to learn the molecular grammars and reaction pre-training to capture fundamental reaction principles, forming a progressive learning paradigm from single- to multi-molecular systems. With prompt-aware downstream fine-tuning, Uni-Mol3 demonstrates exceptional performance in diverse organic reaction tasks and supports multi-task prediction with strong generalizability. Experimental results across 10 datasets spanning 4 downstream tasks show that Uni-Mol3 outperforms existing methods, validating its effectiveness in modeling complex organic reactions. This work not only ushers in an alternative paradigm for multi-molecular computational modeling but also charts a course for intelligent organic reaction by bridging molecular representation with reaction mechanism understanding.

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 1 Pith paper

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. MultiPUFFIN: A Multimodal Domain-Constrained Foundation Model for Molecular Property Prediction of Small Molecules

    cs.LG 2026-03 conditional novelty 6.0 of 10

    MultiPUFFIN claims higher test R² than ChemBERTa-2 on all nine thermophysical properties while using far fewer labeled molecules, with the largest gains on temperature-dependent properties.

Reference graph

Works this paper leans on

58 extracted references · 44 canonical work pages · cited by 1 Pith paper

  1. [1]

    write newline

    " write newline " cite write " FUNCTION editor.postfix editor num.names #1 > "( )" "( )" if FUNCTION editor.trans.postfix editor num.names #1 > "( )" "( )" if FUNCTION trans.postfix translator num.names #1 > "( )" "( )" if FUNCTION authors.editors.reflist.apa5 'field := 'dot := field num.names 'numnames := numnames 'format.num.names := format.num.names na...

  2. [2]

    sn-aps.bst

    FUNCTION identify.aps.version "sn-aps.bst" " [2024/07/19 v1.1 APS bibliography style]" * top ENTRY address author booktitle chapter doi edition editor eid howpublished institution journal key keywords month note number organization pages publisher school series title type url volume year eprint archive archivePrefix primaryClass adsurl adsnote version lab...

    2024

  3. [3]

    write newline

    " write newline "" before.all 'output.state := FUNCTION if.digit duplicate "0" = swap duplicate "1" = swap duplicate "2" = swap duplicate "3" = swap duplicate "4" = swap duplicate "5" = swap duplicate "6" = swap duplicate "7" = swap duplicate "8" = swap "9" = or or or or or or or or or FUNCTION n.separate 't := "" #0 'numnames := t empty not t #-1 #1 subs...

  4. [4]

    , Gao , Z

    botherref Zhou , G. , Gao , Z. , Ding , Q. , Zheng , H. , Xu , H. , Wei , Z. , Zhang , L. , Ke , G. : Uni-mol: A universal 3d molecular representation learning framework (2023) botherref

    work page 2023

  5. [5]

    , Wang , Z

    botherref Ji , X. , Wang , Z. , Gao , Z. , Zheng , H. , Zhang , L. , Ke , G. , et al.: Uni-mol2: Exploring molecular pretraining model at scale. arXiv preprint arXiv:2406.14969 (2024) botherref

    arXiv 2024

  6. [6]

    , Ge , D

    barticle Chu , X.-Q. , Ge , D. , Cui , Y.-Y. , Shen , Z.-L. , Li , C.-J. : Desulfonylation via radical process: recent developments in organic synthesis . Chemical reviews 121 ( 20 ), 12548 -- 12680 ( 2021 ) barticle

    work page 2021

  7. [7]

    , Meng , J

    barticle Ali , R.S.A.E. , Meng , J. , Khan , M.E.I. , Jiang , X. : Machine learning advancements in organic synthesis: A focused exploration of artificial intelligence applications in chemistry . Artificial Intelligence Chemistry 2 ( 1 ), 100049 ( 2024 ) barticle

    work page 2024

  8. [8]

    , Frey , J

    barticle Oliveira , J.C. , Frey , J. , Zhang , S.-Q. , Xu , L.-C. , Li , X. , Li , S.-W. , Hong , X. , Ackermann , L. : When machine learning meets molecular synthesis . Trends in Chemistry 4 ( 10 ), 863 -- 885 ( 2022 ) barticle

    work page 2022

Show all 58 references
  1. [9]

    , Zhao , M

    barticle Dong , J. , Zhao , M. , Liu , Y. , Su , Y. , Zeng , X. : Deep learning in retrosynthesis planning: datasets, models and tools . Briefings in Bioinformatics 23 ( 1 ), 391 ( 2022 ) barticle

    2022

  2. [10]

    , Noh , J

    barticle Chen , S. , Noh , J. , Jang , J. , Kim , S. , Gu , G.H. , Jung , Y. : Reaction templates: Bridging synthesis knowledge and artificial intelligence . Accounts of Chemical Research 57 ( 14 ), 1964 -- 1972 ( 2024 ) barticle

    1964

  3. [11]

    , Renz , P

    barticle Seidl , P. , Renz , P. , Dyubankova , N. , Neves , P. , Verhoeven , J. , Wegner , J.K. , Segler , M. , Hochreiter , S. , Klambauer , G. : Improving few-and zero-shot reaction template prediction using modern hopfield networks . Journal of chemical information and mode...

    2022

  4. [12]

    , Lin , K

    barticle Wang , Z. , Lin , K. , Pei , J. , Lai , L. : Reacon: a template-and cluster-based framework for reaction condition prediction . Chemical Science 16 ( 2 ), 854 -- 866 ( 2025 ) barticle

    2025

  5. [13]

    , Ghosh , A

    barticle Das , M. , Ghosh , A. , Sunoj , R.B. : Advances in machine learning with chemical language models in molecular property and reaction outcome predictions . Journal of Computational Chemistry 45 ( 14 ), 1160 -- 1176 ( 2024 ) barticle

    2024

  6. [14]

    , Zhao , W

    bchapter Li , Y. , Zhao , W. , Dang , B. , Yan , X. , Gao , M. , Wang , W. , Xiao , M. : Research on adverse drug reaction prediction model combining knowledge graph embedding and deep learning . In: 2024 4th International Conference on Machine Learning and Intelligent Systems...

    2024

  7. [15]

    , Song , J

    barticle Zhong , Z. , Song , J. , Feng , Z. , Liu , T. , Jia , L. , Yao , S. , Hou , T. , Song , M. : Recent advances in deep learning for retrosynthesis . Wiley Interdisciplinary Reviews: Computational Molecular Science 14 ( 1 ), 1694 ( 2024 ) barticle

    2024

  8. [16]

    , Laino , T

    barticle Schwaller , P. , Laino , T. , Gaudin , T. , Bolgar , P. , Hunter , C.A. , Bekas , C. , Lee , A.A. : Molecular transformer: a model for uncertainty-calibrated chemical reaction prediction . ACS central science 5 ( 9 ), 1572 -- 1583 ( 2019 ) barticle

    2019

  9. [17]

    , Li , C

    botherref Dai , H. , Li , C. , Coley , C. , Dai , B. , Song , L. : Retrosynthesis prediction with conditional graph logic network. Advances in Neural Information Processing Systems 32 (2019) botherref

    2019

  10. [18]

    , Zhang , Y

    barticle Lu , J. , Zhang , Y. : Unified deep learning model for multitask reaction predictions with explanation . Journal of chemical information and modeling 62 ( 6 ), 1376 -- 1387 ( 2022 ) barticle

    2022

  11. [19]

    , Barzilay , R

    barticle Coley , C.W. , Barzilay , R. , Jaakkola , T.S. , Green , W.H. , Jensen , K.F. : Prediction of organic reaction outcomes using machine learning . ACS central science 3 ( 5 ), 434 -- 443 ( 2017 ) barticle

    2017

  12. [20]

    , Xu , M

    bchapter Shi , C. , Xu , M. , Guo , H. , Zhang , M. , Tang , J. : A graph to graphs framework for retrosynthesis prediction . In: International Conference on Machine Learning , pp. 8818 -- 8827 ( 2020 ). PMLR bchapter

    2020

  13. [21]

    , Estrada , J.G

    barticle Ahneman , D.T. , Estrada , J.G. , Lin , S. , Dreher , S.D. , Doyle , A.G. : Predicting reaction performance in c--n cross-coupling using machine learning . Science 360 ( 6385 ), 186 -- 190 ( 2018 ) barticle

    2018

  14. [22]

    , Schwaller , P

    barticle Probst , D. , Schwaller , P. , Reymond , J.-L. : Reaction classification and yield prediction using the differential reaction fingerprint drfp . Digital discovery 1 ( 2 ), 91 -- 97 ( 2022 ) barticle

    2022

  15. [23]

    , Lowe , D.M

    barticle Schneider , N. , Lowe , D.M. , Sayle , R.A. , Landrum , G.A. : Development of a novel fingerprint for chemical reactions and its application to large-scale reaction classification and similarity . Journal of chemical information and modeling 55 ( 1 ), 39 -- 53 ( 2015 ...

    2015

  16. [24]

    , Vyas , R

    botherref Karthikeyan , M. , Vyas , R. , Karthikeyan , M. , Vyas , R. : Representation, fingerprinting, and modelling of chemical reactions. Practical Chemoinformatics, 317--374 (2014) botherref

    2014

  17. [25]

    , Waller , M.P

    barticle Segler , M.H. , Waller , M.P. : Neural-symbolic machine learning for retrosynthesis and reaction prediction . Chemistry--A European Journal 23 ( 25 ), 5966 -- 5971 ( 2017 ) barticle

    2017

  18. [26]

    , Dwaraknath , S.S

    barticle McDermott , M.J. , Dwaraknath , S.S. , Persson , K.A. : A graph-based network for predicting chemical reaction pathways in solid-state materials synthesis . Nature communications 12 ( 1 ), 3097 ( 2021 ) barticle

    2021

  19. [27]

    , Ierapetritou , M.G

    barticle He , K. , Ierapetritou , M.G. , Androulakis , I.P. : A graph-based approach to developing adaptive representations of complex reaction mechanisms . Combustion and Flame 155 ( 4 ), 585 -- 604 ( 2008 ) barticle

    2008

  20. [28]

    , Coley , C

    botherref Jin , W. , Coley , C. , Barzilay , R. , Jaakkola , T. : Predicting organic reaction outcomes with weisfeiler-lehman network. Advances in neural information processing systems 30 (2017) botherref

    2017

  21. [29]

    , Jin , W

    barticle Coley , C.W. , Jin , W. , Rogers , L. , Jamison , T.F. , Jaakkola , T.S. , Green , W.H. , Barzilay , R. , Jensen , K.F. : A graph-convolutional neural network model for the prediction of chemical reactivity . Chemical science 10 ( 2 ), 370 -- 377 ( 2019 ) barticle

    2019

  22. [30]

    , Dimitriadis , S

    barticle Irwin , R. , Dimitriadis , S. , He , J. , Bjerrum , E.J. : Chemformer: a pre-trained transformer for computational chemistry . Machine Learning: Science and Technology 3 ( 1 ), 015022 ( 2022 ) barticle

    2022

  23. [31]

    , Zhang , Z

    barticle Jiang , S. , Zhang , Z. , Zhao , H. , Li , J. , Yang , Y. , Lu , B.-L. , Xia , N. : When smiles smiles, practicality judgment and yield prediction of chemical reaction via deep chemical language processing . IEEE Access 9 , 85071 -- 85083 ( 2021 ) barticle

    2021

  24. [32]

    , Vaucher , A.C

    barticle Schwaller , P. , Vaucher , A.C. , Laino , T. , Reymond , J.-L. : Prediction of chemical reaction yields using deep learning . Machine learning: science and technology 2 ( 1 ), 015016 ( 2021 ) barticle

    2021

  25. [33]

    found in translation

    barticle Schwaller , P. , Gaudin , T. , Lanyi , D. , Bekas , C. , Laino , T. : “found in translation”: predicting outcomes of complex organic chemistry reactions using neural sequence-to-sequence models . Chemical science 9 ( 28 ), 6091 -- 6098 ( 2018 ) barticle

    2018

  26. [34]

    , Godin , G

    bchapter Karpov , P. , Godin , G. , Tetko , I.V. : A transformer model for retrosynthesis . In: International Conference on Artificial Neural Networks , pp. 817 -- 830 ( 2019 ). Springer bchapter

    2019

  27. [35]

    : Smiles, a chemical language and information system

    barticle Weininger , D. : Smiles, a chemical language and information system. 1. introduction to methodology and encoding rules . Journal of chemical information and computer sciences 28 ( 1 ), 31 -- 36 ( 1988 ) barticle

    1988

  28. [36]

    : Introduction to Stereochemistry

    bbook Mislow , K. : Introduction to Stereochemistry . Courier Corporation , ??? ( 2012 ) bbook

    2012

  29. [37]

    , Poppe , L

    bbook N \'o gr \'a di , M. , Poppe , L. , Nagy , J. , Horny \'a nszky , G. , Boros , Z. : Stereochemistry and Stereoselective Synthesis: An Introduction . John Wiley & Sons , ??? ( 2016 ) bbook

    2016

  30. [38]

    , Flamm , C

    bchapter Andersen , J.L. , Flamm , C. , Merkle , D. , Stadler , P.F. : Chemical graph transformation with stereo-information . In: Graph Transformation: 10th International Conference, ICGT 2017, Held as Part of STAF 2017, Marburg, Germany, July 18-19, 2017, Proceedings 10 , pp...

    2017

  31. [39]

    , Shao , Y

    bchapter Cao , H. , Shao , Y. , Liu , Z. , Liu , Z. , Tang , X. , Yao , Y. , Li , Y. : PRESTO : Progressive pretraining enhances synthetic chemistry outcomes . In: Al-Onaizan , Y. , Bansal , M. , Chen , Y.-N. (eds.) Findings of the Association for Computational Linguistics: EM...

    2024

  32. [40]

    , Minnen , D

    bchapter Mentzer , F. , Minnen , D. , Agustsson , E. , Tschannen , M. : Finite scalar quantization: VQ - VAE made simple . In: The Twelfth International Conference on Learning Representations ( 2024 ). https://openreview.net/forum?id=8ishA3LxN8 bchapter

    2024

  33. [41]

    , Irwin , J.J

    barticle Sterling , T. , Irwin , J.J. : Zinc 15--ligand discovery for everyone . Journal of chemical information and modeling 55 ( 11 ), 2324 -- 2337 ( 2015 ) barticle

    2015

  34. [42]

    , Bellis , L.J

    barticle Gaulton , A. , Bellis , L.J. , Bento , A.P. , Chambers , J. , Davies , M. , Hersey , A. , Light , Y. , McGlinchey , S. , Michalovich , D. , Al-Lazikani , B. , : Chembl: a large-scale bioactivity database for drug discovery . Nucleic acids research 40 ( D1 ), 1100 -- 1...

    2012

  35. [43]

    , Ramsundar , B

    barticle Liu , B. , Ramsundar , B. , Kawthekar , P. , Shi , J. , Gomes , J. , Luu Nguyen , Q. , Ho , S. , Sloane , J. , Wender , P. , Pande , V. : Retrosynthetic reaction prediction using neural sequence-to-sequence models . ACS central science 3 ( 10 ), 1103 -- 1113 ( 2017 ) barticle

    2017

  36. [44]

    , Hsieh , C.-Y

    barticle Wang , X. , Hsieh , C.-Y. , Yin , X. , Wang , J. , Li , Y. , Deng , Y. , Jiang , D. , Wu , Z. , Du , H. , Chen , H. , : Generic interpretable reaction condition predictions with open reaction condition datasets and unsupervised learning of reaction center . Research 6...

    2023

  37. [45]

    , : Binary codes capable of correcting deletions, insertions, and reversals

    bchapter Levenshtein , V.I. , : Binary codes capable of correcting deletions, insertions, and reversals . In: Soviet Physics Doklady , vol. 10 , pp. 707 -- 710 ( 1966 ). Soviet Union bchapter

    1966

  38. [46]

    , Landrum , G.A

    barticle Riniker , S. , Landrum , G.A. : Better informed distance geometry: using what we know to improve conformation generation . Journal of chemical information and modeling 55 ( 12 ), 2562 -- 2574 ( 2015 ) barticle

    2015

  39. [47]

    : Merck molecular force field

    barticle Halgren , T.A. : Merck molecular force field. i. basis, form, scope, parameterization, and performance of mmff94 . Journal of computational chemistry 17 ( 5-6 ), 490 -- 519 ( 1996 ) barticle

    1996

  40. [48]

    , Shazeer , N

    barticle Raffel , C. , Shazeer , N. , Roberts , A. , Lee , K. , Narang , S. , Matena , M. , Zhou , Y. , Li , W. , Liu , P.J. : Exploring the limits of transfer learning with a unified text-to-text transformer . Journal of machine learning research 21 ( 140 ), 1 -- 67 ( 2020 ) barticle

    2020

  41. [49]

    sn-basic.bst

    FUNCTION identify.basic.version "sn-basic.bst" " [2024/07/19 v1.1 bibliography style]" * top ENTRY address archive author booktitle chapter doi edition editor eid eprint howpublished institution journal key keywords month note number organization pages publisher school series ...

    2024

  42. [50]

    write newline

    " write newline "" before.all 'output.state := FUNCTION add.period duplicate empty 'skip "." * add.blank if FUNCTION if.digit duplicate "0" = swap duplicate "1" = swap duplicate "2" = swap duplicate "3" = swap duplicate "4" = swap duplicate "5" = swap duplicate "6" = swap dupl...

  43. [51]

    write newline

    " write newline "" before.all 'output.state := FUNCTION output.doi doi empty skip "doi:" doi * "" * output if FUNCTION format.archive archivePrefix empty "" archivePrefix ":" * if FUNCTION format.primaryClass primaryClass empty "" " [" primaryClass * "] " * if FUNCTION format....

  44. [52]

    write newline

    " write newline "" before.all 'output.state := FUNCTION string.to.integer 't := t text.length 'k := #1 'char.num := t char.num #1 substring 's := s is.num s "." = or char.num k = not and char.num #1 + 'char.num := while char.num #1 - 'char.num := t #1 char.num substring FUNCTI...

  45. [53]

    write newline

    " write newline "" before.all 'output.state := FUNCTION string.to.integer 't := t text.length 'k := #1 'char.num := t char.num #1 substring 's := s is.num s "." = or char.num k = not and char.num #1 + 'char.num := while char.num #1 - 'char.num := t #1 char.num substring FUNCTI...

  46. [54]

    sn-nature.bst

    FUNCTION identify.nature.version "sn-nature.bst" " [2024/07/19 v1.1 bibliography style]" * top ENTRY address archive author booktitle chapter edition editor eprint howpublished institution journal key keywords month note number organization pages publisher school series title ...

    2024

  47. [55]

    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 gl...

  48. [56]

    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 gl...

  49. [57]

    sn-vancouver-num.bst

    FUNCTION identify.vancouver.version "sn-vancouver-num.bst" " [2024/07/19 v1.1 Vancouver bibliography style]" * top ENTRY address assignee author booktitle chapter cartographer day edition editor howpublished institution inventor journal key keywords month note number organizat...

    2024

  50. [58]

    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 gl...

Pith tools

Reviewed August 6, 2026 · model on record in the stance chip above.