Notice of Pre-AIA or AIA Status
The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA .
Information Disclosure Statement
The filed information disclosure statement (IDS) is in compliance with the provisions of 37 CFR 1.97. Accordingly, the information disclosure statement is being considered by the examiner.
Claim Rejections - 35 USC § 103
In the event the determination of the status of the application as subject to AIA 35 U.S.C. 102 and 103 (or as subject to pre-AIA 35 U.S.C. 102 and 103) is incorrect, any correction of the statutory basis (i.e., changing from AIA to pre-AIA ) for the rejection will not be considered a new ground of rejection if the prior art relied upon, and the rationale supporting the rejection, would be the same under either status.
The following is a quotation of 35 U.S.C. 103 which forms the basis for all obviousness rejections set forth in this Office action:
A patent for a claimed invention may not be obtained, notwithstanding that the claimed invention is not identically disclosed as set forth in section 102, if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains. Patentability shall not be negated by the manner in which the invention was made.
Claims 1-2, 4-9, 11-16, and 18-20 are rejected under 35 U.S.C. 103 as being unpatentable over Schriber (US 2020/0193718) in view of Majumdar (US 2020/0004752).
As per claim 1, Schriber teaches determining a path language for embedding paths by one or more robots through one or more embedded environments representing a corresponding one or more physical environments, the one or more embedded environments including a plurality of environment parts representing elements of the one or more physical environments, the path language including a plurality of nouns corresponding with the plurality of environment parts, the path language including a plurality of verbs corresponding with actions affecting the one or more robots within the one or more physical environments, the path language including a grammar defining construction of one or more sentence types based on the nouns and the verbs (during development of a virtual real (VR) experience, a script, in text format, may be entered by a user, a frontend analytics engine may be used to parse the script. The frontend analytics engine may be programmed with natural language processing functionality such that it can analyze the text of a script and determine the existence of meaningful language, such as language indicative of characters, actions, interactions between characters, dialog, etc. For example, frontend analytics engine may extract metadata indicative of a character, e.g., based on text determined or known to be a name. Frontend analytics engine may extract metadata indicative of an action, e.g., based on text indicative of action, such as verbs. Frontend analytics engine may extract metadata indicative of location, e.g., based on known names of locations (geographical and/or set location) or other textual location information found in the scrip, [0022]- [0027]);
Schriber may not explicitly disclose determining via a processor a plurality of path language embeddings corresponding with a plurality of paths through the one or more embedded environments, each of the plurality of path language embeddings including a respective one or more sentences determined in accordance with the grammar; determining a novel path for a designated robot based on a novel path embedding generated by a path large language model, the path large language model including a foundational large language model and a fine-tuning layer tuned based on the plurality of paths; and transmitting an instruction to the robot to traverse all or a portion of the novel path.
Majumdar in the same field of endeavor teaches determining a plurality of paths language embeddings corresponding with a plurality of paths through the one or more embedded environments, each of the plurality of path language embeddings including a respective one or more sentences determined in accordance with the grammar (Fig. 2, [0025], [0032]- [0036], [0115], and [0148], wherein determined, a plurality of paths between nodes/vertices of the network/graph represented as weighted edges); determining a novel path for a designated robot based on a novel path embedding generated by a path large language model, the path large language model including a foundational large language model and a fine-tuning layer tuned based on the plurality of paths (Majumdar, [0011], backpropagation is the most common ANN learning method as it uses error/loss function to fine tune the network. See also, [0080], wherein redundant paths are removed); and transmitting an instruction to the robot to traverse all or a portion of the novel path (Majumdar, [0086], wherein the best ranking vector are used for information retrieval (by search engines/robots)).
Therefore, it would have been obvious at the time the application was filed to use Majumdar’s features with the system of Schriber, in order to improve accuracy and enhance efficiency of automated tasks.
As per claim 2, Schriber does not explicitly disclose wherein the fine-tuning layer is tuned via reinforcement learning based on a loss function determined based on a plurality discrete Feynman path integrals computed over the plurality of paths in the one or more embedded environments. Majumdar in the same field of endeavor teaches computing discrete Feynman path integrals to traverse the plurality of paths of the network, wherein the propagation operator is treated as a summation/integral over all possible states to rank and determine a novel path representing an input query ([0011], [0025], [0073], and [0074]). Therefore, it would have been obvious at the time the application was filed to use Majumdar’s discrete Feynman path integrals to traverse feature with the system of Schriber, in order to offer practical solutions for problems in the analysis of complex systems like neural networks.
As per claim 4, Schriber does not explicitly disclose wherein the loss function penalizes paths associated with more negative outcome values. Majumdar in the same field of endeavor teaches wherein the loss function penalizes paths associated with more negative outcome values ([0011], [0043], [0080], applying redundant path erasure function). Therefore, it would have been obvious at the time the application was filed to use Majumdar’s above feature with the system of Schriber, in order to provide natural language models with more accurate and appropriate outputs.
As per claim 5, Schriber does not explicitly disclose determining a plurality of outcome values corresponding with the plurality of paths by retrieving path outcome information from a database system. Majumdar in the same field of endeavor teaches determining a plurality of outcome values corresponding with the plurality of paths by retrieving path outcome information from a database system ([0171], Fig. 2, [0043]- [0046] and claim 1). Therefore, it would have been obvious at the time the application was filed to use Majumdar’s above feature with the system of Schriber, in order to provide natural language models with more accurate and appropriate outputs.
As per claim 6, Schriber teaches wherein a designated noun of the plurality of nouns includes a designated identifier that uniquely identifies a corresponding designated environment part of the plurality of environment parts ([0027], wherein said, frontend analytics engine may extract metadata indicative of location, e.g., based on known names of locations (geographical and/or set location) or other textual location information found in the script).
As per claim 7, Schriber teaches wherein the grammar includes a modifier, wherein a designated sentence type of the one or more sentence types includes the modifier, and wherein the modifier identifies a length of time that an associated action occurred ([0037], wherein one or more of frontend and backend analytics engines may adjust any event timing or timing checks for consistency among different aspects of a scene).
As per claims 8, 9, and 11-14, Schriber teaches a computer readable medium ([0066]). The remaining steps are rejected under the same rationale as applied to the method steps of rejected claims 1-2 and 4-7.
As per claims 15-16 and 18-20, system claims 15-16 and 18-20 and method claims 1-2, 4, 6, and 7 are related as apparatus and the method of using same, with each claimed element's function corresponding to the claimed method step. Accordingly claims 15-16 and 18-20 are similarly rejected under the same rationale as applied above with respect to method claims 1-2, 4, 6, and 7. Furthermore, Schriber teaches a processor, memory, and a communication interface, as claimed (Fig. 5).
Claims 3, 10, and 17 are rejected under 35 U.S.C. 103 as being unpatentable over Schriber (US 2020/0193718) in view of Majumdar (US 2020/0004752), and further in view of Mathur (2024/0412226).
As per claim 3, 10, and 17, Schriber in view of Majumdar teaches a method, computer readable medium, and system as described by independent claims 1, 8, and 15. Schriber in view of Majumdar does not explicitly disclose wherein paths that occur more frequently contribute more weight to the plurality of discrete Feynman path integrals. Mathur (2024/0412226) in the same field of endeavor teaches determining and evaluating the extracted path of actions based on a path frequency metric ([0005]). The term “action path” refers to a sequence of terms included in a response from the language model ([0017], [0036]). Therefore, it would have been obvious at the time the application was filed to use Mathur’s frequency of occurrence metric with the system of Schriber in view of Majumdar, in order to provide straightforward and simplified natural language processing compared to more complex machine learning models.
Conclusion
The prior art made of record and not relied upon is considered pertinent to applicant's disclosure. See PTO-892.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to ABDELALI SERROU whose telephone number is (571)272-7638. The examiner can normally be reached M-F 9 Am - 5 PM.
Examiner interviews are available via telephone, in-person, and video conferencing using a USPTO supplied web-based collaboration tool. To schedule an interview, applicant is encouraged to use the USPTO Automated Interview Request (AIR) at http://www.uspto.gov/interviewpractice.
If attempts to reach the examiner by telephone are unsuccessful, the examiner’s supervisor, Pierre-Louis Desir can be reached at 571-272-7799. The fax phone number for the organization where this application or proceeding is assigned is 571-273-8300.
Information regarding the status of published or unpublished applications may be obtained from Patent Center. Unpublished application information in Patent Center is available to registered users. To file and manage patent submissions in Patent Center, visit: https://patentcenter.uspto.gov. Visit https://www.uspto.gov/patents/apply/patent-center for more information about Patent Center and https://www.uspto.gov/patents/docx for information about filing in DOCX format. For additional questions, contact the Electronic Business Center (EBC) at 866-217-9197 (toll-free). If you would like assistance from a USPTO Customer Service Representative, call 800-786-9199 (IN USA OR CANADA) or 571-272-1000.
/ABDELALI SERROU/ Primary Examiner, Art Unit 2659