Neural Message Passing for Quantum Chemistry

Neural Message Passing for Quantum Chemistry Justin Gilmer Samuel S. Schoenholz Patrick F. Riley Oriol Vinyals George E. Dahl

Tag layer of the DHSS Meta-Tagging Project. The article text is reproduced from the staged copy named in sourceDesc and is not altered; the tags are held separately in standOff so that overlapping and nested spans are preserved.

Neural Message Passing for Quantum Chemistry 2017 ICML / arXiv 1704.01212 gilmer2017-message-passing gilmer2017-message-passing.txt, the project's staged copy b3bbdb58f69b78b927542774a719ddd66e6eeced arxiv-open chemistry

the paper's own words

the tag layer; a sentence written by the tagger about the paper, always anchored to a quotation but not itself a quotation

0 spans located by searching the stored text for the quote's exact words, allowing any whitespace between them method Quotation withheld because this record is not cleared for redistribution. molecular property prediction from graphs Quotation withheld because this record is not cleared for redistribution. system-or-tool-building system-or-tool-building Quotation withheld because this record is not cleared for redistribution. mechanisms message passing on the molecular graph (atoms as nodes, bonds as edges): iteratively passing and aggregating messages between atoms Quotation withheld because this record is not cleared for redistribution. the models fit the DFT approximation, itself an approximation of true quantum properties what this limits: every accuracy claim: the target is THE DFT APPROXIMATION, so 'chemical accuracy' is accuracy to a model rather than to the quantum properties themselves Quotation withheld because this record is not cleared for redistribution. The towers variant improves scalability but is still insufficient for much larger graphs. what this limits: the scalability claim: THE TOWERS VARIATION MAKES THESE MODELS MORE SCALABLE, BUT ADDITIONAL IMPROVEMENTS WILL BE NEEDED TO SCALE TO MUCH LARGER GRAPHS Quotation withheld because this record is not cleared for redistribution. Generalising to larger molecules is hard precisely when spatial information is used. what this limits: the spatial-information variants specifically: GENERALIZING TO LARGER MOLECULE SIZES SEEMS PARTICULARLY CHALLENGING WHEN USING SPATIAL INFORMATION Quotation withheld because this record is not cleared for redistribution. Combining the towers variant with the edge-network message function did not work. what this limits: the towers-plus-edge-network combination: UNFORTUNATELY, OUR ATTEMPTS SO FAR AT COMBINING THE TOWERS AND EDGE NETWORK MESSAGE FUNCTION HAVE FAILED TO FURTHER IMPROVE PERFORMANCE Quotation withheld because this record is not cleared for redistribution. Some ablation experiments used a partial-charge input feature that would not be available in a real application. what this limits: the ablations using partial charge, which no deployment could reproduce: THIS FEATURE IS AN OUTPUT OF THE DFT CALCULATION AND THUS COULD NOT BE USED IN AN APPLIED SETTING Quotation withheld because this record is not cleared for redistribution. Accuracy is measured against DFT approximations because experimentally true targets do not exist for the full dataset. what this limits: every accuracy figure on this benchmark, which is agreement with a computation rather than with measurement: THE DATASET CONTAINING THE TRUE TARGETS ON ALL 134k MOLECULES DOES NOT CURRENTLY EXIST Quotation withheld because this record is not cleared for redistribution. The paper's headline term is redefined to mean error against its own labels rather than against nature, and it says so in one sentence. what this limits: the claim that this model reaches chemical accuracy: the authors state that in the rest of the paper the phrase means accuracy with respect to their available ground truth labels, not the community's target error relative to nature Quotation withheld because this record is not cleared for redistribution. Two different error standards are in play and the paper orders them. what this limits: a comparison of the errors reported here against either standard: the paper states that Quotation withheld Quotation withheld because this record is not cleared for redistribution. Eight existing models are re-described in the authors' own framework, with the derivations placed outside the paper. what this limits: the claim that this framework unifies prior work: at least eight models from the literature are described in it, and the exact expressions and the derivation of that reformulation are in the supplementary material rather than the paper Quotation withheld because this record is not cleared for redistribution. graph neural networks / deep learning, applied to predict quantum-chemical properties of molecules Quotation withheld because this record is not cleared for redistribution. the model family evaluated on all 13 targets of the public QM9 benchmark eliminates: a chemistry result earned on molecules of the authors' own choosing - QM9 is fixed, public and shared with the methods being compared against · outcome (stated by the tagger, no quotation): held Quotation withheld because this record is not cleared for redistribution. chemical accuracy chemical accuracy Quotation withheld because this record is not cleared for redistribution. DFT error DFT error Quotation withheld because this record is not cleared for redistribution. Message Passing Neural Networks Message Passing Neural Networks Quotation withheld because this record is not cleared for redistribution. ground truth labels ground truth labels Quotation withheld because this record is not cleared for redistribution. QM9 QM9 Quotation withheld because this record is not cleared for redistribution. 'chemical accuracy' - TWO REFERENCE POINTS, AND THE PAPER PICKS THE NEARER ONE: the chemistry community's target error is relative to nature, DFT error is the gap between the simulation and nature, and this paper states that its own use of the phrase means error against its available labels - so the headline measures agreement with a simulation, not with a molecule Quotation withheld because this record is not cleared for redistribution. MPNN - A NAME GIVEN TO EIGHT EXISTING MODELS AFTER THE FACT: the framework is introduced as a reformulation of models already in the literature under one label, which is the inverse of the usual naming failure - here a paper renames its field rather than being renamed by it Quotation withheld because this record is not cleared for redistribution. the prediction target - A MODEL OF A MODEL: what the network predicts is the outcome of a computationally expensive DFT calculation, so the object being learned is itself a simulation of the quantity of interest rather than the quantity Quotation withheld because this record is not cleared for redistribution. builds-on → graph neural networks (invariant to molecular symmetries) Quotation withheld because this record is not cleared for redistribution. builds-on → Gated Graph Neural Networks (GG-NN), Li et al. (2016) Quotation withheld because this record is not cleared for redistribution. builds-on → the set2set model, Vinyals et al. (2015) Quotation withheld because this record is not cleared for redistribution. extends → Interaction Networks, Battaglia et al. (2016) Quotation withheld because this record is not cleared for redistribution. chemical accuracy —operationalised-as→ ground truth labels chemical accuracy ground truth labels strength of the relation: 1 Quotation withheld because this record is not cleared for redistribution. chemical accuracy —distinct-from→ DFT error chemical accuracy DFT error strength of the relation: 1 Quotation withheld because this record is not cleared for redistribution. 130k — size of the QM9 benchmark with 13 properties each Quotation withheld because this record is not cleared for redistribution. 300k — MPNN inference speedup over DFT simulation Quotation withheld because this record is not cleared for redistribution. 11 — targets predicted within chemical accuracy by the best MPNN Quotation withheld because this record is not cleared for redistribution. 5 — targets hit using molecular topology alone about this quote: marked quote-not-in-current-staged-text - today's merge gate would REFUSE this quote: Paper.locate() does not find it in the current staged text, and no disclosure on the row explains why. Update, 15 September 2026: CLEARED evidence status='quote-not-in-current-staged-text' - the quote was replaced (EXTEND) and the new quote locates exactly once in the current staged text. Quotation withheld because this record is not cleared for redistribution. 8 — prior models unified under the MPNN framework Quotation withheld because this record is not cleared for redistribution. MPNN variants achieve a new state of the art on the QM9 dataset Quotation withheld because this record is not cleared for redistribution. neural inference is far cheaper than the DFT simulation it approximates Quotation withheld because this record is not cleared for redistribution. the readout must be permutation-invariant for graph-isomorphism invariance Quotation withheld because this record is not cleared for redistribution. Duvenaud-style message passing cannot capture edge-node correlations Quotation withheld because this record is not cleared for redistribution. hidden states hidden states Quotation withheld because this record is not cleared for redistribution. time steps time steps Quotation withheld because this record is not cleared for redistribution. node features node features Quotation withheld because this record is not cleared for redistribution. edge features edge features Quotation withheld because this record is not cleared for redistribution. update functions update functions Quotation withheld because this record is not cleared for redistribution. readout functions readout functions Quotation withheld because this record is not cleared for redistribution. Message Passing Neural Networks (MPNNs) Message Passing Neural Networks (MPNNs) Quotation withheld because this record is not cleared for redistribution. chemical accuracy chemical accuracy Quotation withheld because this record is not cleared for redistribution. DFT error DFT error Quotation withheld because this record is not cleared for redistribution. readout phase readout phase Quotation withheld because this record is not cleared for redistribution. sample-size sample-size Quotation withheld because this record is not cleared for redistribution.

The full text of this article is not redistributable, so it is not included here. The evidence sentences are withheld with it; the tag layer and its grounding status remain. The article itself is at https://arxiv.org/abs/1704.01212.