# US 12,769,790 B2 · Neural networks to indicate data dependencies > Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt - Number: US 12,769,790 B2 - Document ID: US12769790B2 - Status: Granted - Kind: B2 - Application number: 18/098,015 - Filed: 2023-01-17 - Published: 2024-04-18 - Granted: 2026-10-20 - Earliest claimed: 2022-05-19 - Assignees: NVIDIA Corporation - Applicants: NVIDIA Corporation - Inventors: Marc Teva Law; James Robert Lucas - Web page: https://patentgrep.com/patent/US12769790B2 ## CPC - G06F 16/9024: {Graphs; Linked lists (G06F16/9027 takes precedence)} - G06N 3/09: Supervised learning - G06N 3/044: Recurrent networks, e.g. Hopfield networks - G06N 3/0464: Convolutional networks [CNN, ConvNet] - G06N 3/063: using electronic means - G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N) - G06N: COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS ## Abstract Apparatuses, systems, and techniques to indicate data dependencies. In at least one embodiment, one or more neural networks are used to generate one or more indicators of one or more data dependencies and one or more indicators of direction of the one or more data dependencies. ## Claims ### Claim 1 A processor, comprising: one or more circuits to: receive data input for a directed graph; use one or more neural networks to generate a spacetime embedding of the directed graph, based on the data input for the directed graph, the spacetime embedding representing one or more data dependencies and direction of the one or more data dependencies within the directed graph as one or more causal connections between events in a spacetime, wherein to generate the spacetime embedding the one or more neural networks are to restrict a connected pair of nodes of the directed graph to be represented as causally-connected events within a same convex normal neighborhood of the spacetime such that the spacetime embedding is restricted to events connected on a time-directed curve within a threshold distance of one another in the same convex normal neighborhood of the spacetime; and perform inferencing using one or more neural networks and based at least on the spacetime embedding. ### Claim 2 (depends on 1) The processor of claim 1, wherein the one or more causal connections represent respective directions of the one or more data dependencies as one or more chronological orderings in the spacetime. ### Claim 3 (depends on 1) The processor of claim 1, wherein the one or more data dependencies include one or more edges of the directed graph and the direction of the one or more data dependencies include a direction of each edge of the one or more edges. ### Claim 4 (depends on 1) The processor of claim 1, wherein the spacetime embedding of the directed graph is to be generated by the one or more neural networks based, at least in part, on whether events in the spacetime, corresponding to nodes of the directed graph, exist within a shared convex normal neighborhood of the spacetime. ### Claim 5 (depends on 1) The processor of claim 1, wherein the spacetime embedding of the directed graph is to be generated by the one or more neural networks based, at least in part, on a sign of a time function corresponding to events in the spacetime. ### Claim 6 (depends on 1) The processor of claim 1, wherein to train, or to perform inferencing using the one or more neural networks is based on one or more distances between events according to coordinates of the events in the spacetime of the spacetime embedding. ### Claim 7 (depends on 1) The processor of claim 1, wherein the one or more causal connections represent respective directions of the one or more data dependencies at least by representing a direction of each data dependency of the one or more data dependencies as a direction of time in the spacetime. ### Claim 8 A system, comprising memory to store executable instructions that, if executed by one or more processors, cause the system to: receive data input for a directed graph; use one or more neural networks to generate a spacetime embedding of the directed graph, based on the data input for the directed graph, the spacetime embedding representing one or more data dependencies and direction of the one or more data dependencies within the directed graph as one or more causal connections between events in a spacetime, wherein to generate the spacetime embedding the one or more neural networks are to restrict a connected pair of nodes of the directed graph to be represented as causally-connected events within a same convex normal neighborhood of the spacetime such that the spacetime embedding is restricted to events connected on a time-directed curve within a threshold distance of one another in the same convex normal neighborhood of the spacetime; and perform inferencing using one or more neural networks and based at least on the spacetime embedding. ### Claim 9 (depends on 8) The system of claim 8, wherein the one or more neural networks are to generate the spacetime embedding representing the directed graph in a pseudo-Riemannian manifold of the spacetime. ### Claim 10 (depends on 8) The system of claim 8, wherein the spacetime embedding of the directed graph represents one or more cycles of the directed graph and a direction corresponding to each cycle of the one or more cycles. ### Claim 11 (depends on 8) The system of claim 8, wherein the spacetime embedding of the directed graph is to be generated by the one or more neural networks based, at least in part, on whether events in the spacetime, corresponding to nodes of the directed graph, exist within the convex normal neighborhood of one another. ### Claim 12 (depends on 8) The system of claim 8, wherein the spacetime embedding of the directed graph is to be generated by the one or more neural networks based, at least in part, on a sign of a time function corresponding to a causal ordering of events in the spacetime. ### Claim 13 (depends on 8) The system of claim 8, wherein the one or more neural networks are to generate the spacetime embedding representing the one or more data dependencies at least by representing the one or more data dependencies as one or more non-spacelike geodesics between events in the spacetime. ### Claim 14 (depends on 8) The system of claim 8, wherein the one or more neural networks are to generate the spacetime embedding representing direction of the one or more data dependencies at least by representing a direction of each data dependency of the one or more data dependencies as a past time orientation or a future time orientation in the spacetime. ### Claim 15 A method, comprising: receiving data input for a directed graph; using one or more neural networks to generate a spacetime embedding of the directed graph, based on the data input for the directed graph, the spacetime embedding representing one or more data dependencies and direction of the one or more data dependencies within the directed graph as one or more causal connections between events in a spacetime, wherein to generate the spacetime embedding the one or more neural networks are to restrict a connected pair of nodes of the directed graph to be represented as causally-connected events within a same convex normal neighborhood of the spacetime such that the spacetime embedding is restricted to events connected on a time-directed curve within a threshold distance of one another in the same convex normal neighborhood of the spacetime; and performing inferencing using one or more neural networks and based at least on the spacetime embedding. ### Claim 16 (depends on 15) The method of claim 15, wherein using the one or more neural networks to generate the spacetime embedding of the directed graph comprises representing the one or more data dependencies in a Lorentz manifold of the spacetime, and wherein he one or more data dependencies include one or more directed cycles of the directed graph and the direction of the one or more data dependencies include a direction of each directed cycle of the one or more directed cycles. ### Claim 17 (depends on 15) The method of claim 15, wherein the spacetime embedding of the directed graph is to be generated by the one or more neural networks based, at least in part, on whether events in the spacetime, corresponding to nodes of the directed graph, exist within an open convex normal neighborhood of one another. ### Claim 18 (depends on 15) The method of claim 15, wherein the spacetime embedding of the directed graph is to be generated by the one or more neural networks based, at least in part, on a sign of a time function corresponding to a chronological ordering of events in the spacetime. ### Claim 19 (depends on 15) The method of claim 15, wherein the one or more neural networks are to generate the spacetime embedding of the directed graph at least by restricting the one or more data dependencies to correspond to one or more timelike geodesics between events in the spacetime. ### Claim 20 (depends on 15) The method of claim 15, wherein the one or more neural networks are to generate the spacetime embedding of the directed graph at least by restricting a direction of each data dependency of the one or more data dependencies to correspond to a future time orientation in the spacetime.