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 .
Compact Prosecution
Examiner would like to Propose Examiners interview to discuss possible amendments to move prosecution forward.
Response to Amendment
Applicant on page 6 in the Remarks commented that the office action is not complete and clear because it fails to identify any claims which the examiner judges are presented to be allowable and cited MPEP 707.07(d).
However, MPEP 707.07(d) talks about: (Language To be Used in Rejecting Claims)
Where a claim is refused for any reason relating to the merits thereof it should be "rejected" and the ground of rejection fully and clearly stated, and the word "reject" must be used. The examiner should designate the statutory basis for any ground of rejection by express reference to a section of 35 U.S.C. in the opening sentence of each ground of rejection. Claims should not be grouped together in a common rejection unless that rejection is equally applicable to all claims in the group.
This has nothing to do with the incompletes of the office action. Examiner therefore seeks clarifications for the argument.
Response to Arguments
Applicant’s arguments with respect to claim(s) 1-18 have been considered but are not persuasive.
Applicant argues that the combination of the references fails to address the limitation “generating a respective rationale to determine a respective output and generating a respective context, based on the respective rationale.
In reply Examiner respectfully disagree because Brown discloses generating a respective rationale to determine a respective output; (Section 0031, lines 1-3 Fig. 1 which is a system that depicts analyzing input or queries and generate answers by ranking the answers are based on a logical architecture)
NB: this means that the answers (respective output) are generated based on logical or rationale architecture- Also see Section 0038.
Claim Rejections - 35 USC § 103
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 1-18 are rejected under 35 U.S.C. 103 as being unpatentable over Brown 2012/0078891 in view of Ferrucci 20200034428
Claim 1, Brown discloses a method comprising generating, using a multimodal model, (Model 1, 2, ---- Model N shown in Fig. 7) a response to a query; (Answer A1, Answer Z1 and Answer Z2 in fig. 7)
iteratively determining, (Section 0007, lines 6-8 “ranging the questions” from broad to narrow shows the iteratively parsing the input) in response to the query whether one or more subsequent outputs of the multimodal model satisfies the query; (Section 0012, lines 4-5 based on the search select or identify candidate answers to the input query)
generating a respective rationale to determine a respective output; (Section 0031, lines 1-3 Fig. 1 which is a system that depicts analyzing input or queries and generate answers by ranking the answers are based on a logical architecture)
NB: this means that the answers (respective output) are generated based on logical or rationale architecture- Also see Section 0038.
generating a respective associated with a respective output of the multimodal model; (Section 0031, lines 7-9 refers to a query (and its context)) and
determining, by the multimodal model based on the respective, whether the respective subsequent output satisfies the query. (Section 0033, lines 7-9 query response is generated based one or more of multiple modalities including text, audio etc.)
It is not clear how Brown discloses query in context of the multimodal model.
Ferrucci discloses model in some common context. (Section 0093, lines 2 “Task related and generally context dependent”)
Therefore it will be obvious to one ordinary skilled in the art before the effective filing date of the claimed invention to include the teaching of analyzing inputs or queries in context. The motivation is that it makes the answers very accurate.
Claim 2, Brown in view of Ferrucci discloses that the method of claim 1, further comprising obtaining at least a portion of one or more data records based on the query; (Brown: Section 0038, lines 17-20, Brown: decomposing an input query into it different components such as grammatical , lexical and syntactic and predicate structure and these components reads on the data records) generating an initial prompt based on the at least a portion of the one or more data records; (Brown: Section 0038, lines 7-9- thus a string and an implicit context or inner questions reads on the initial prompt) and wherein the multimodal model determines whether the respective subsequent output of the multimodal model satisfies the query based on the prompt. (Brown, Section 0039, lines 18-20- thus answer (subsequent output) to an inner question reads on the initial prompt as well).
Claim 3, Brown in view of Ferrucci discloses the method of claim 2, further comprising obtaining at least another portion of the one or more data records based on the respective subsequent output of the multimodal model; (Brown, Section 0038, lines 18-19- thus the Question Classifier block reads on at least another portion of the plurality of data records) generating a subsequent prompt based on the respective subsequent output of the multimodal model and the respective context; (Brown, Section 0039, lines 14-15- question about the target answer reads on the subsequent prompt) and wherein the multimodal model determines whether the respective subsequent output satisfies the query based on the subsequent prompt. (Brown, Section 0039, lines 21-23- thus uniquely identifying the answer to the query)_
Claim 4, Brown in view of Ferrucci discloses the method of claim 3, wherein each portion of the at least a portion of the one or more data records is associated with an embedding value, (Brown, Section 0059, lines 15-17 thus values 0.46, 0.48 reads on the values) and wherein the at least a portion of the one or more data records is obtained based on the embedding value and an embedding index. (Brown: Section 0043, lines 3-7 thus providing an index into the tree)
Claim 5, Brown in view of Ferrucci discloses the method of claim 4, further comprising obtaining, by a retriever module, the at least a portion of the one or more data records, and wherein the retriever module provides the at least a portion of the one or more data records to the multimodal model. (Brown, Section 0060, lines 5-8 search engine which is performing retrieval function)
Claim 6, Brown in view of Ferrucci discloses the method of claim 5, wherein the retriever module comprises a multimodal machine learning model. (Brown: Section 0002, lines 2 retrieval to question answering, See Section 0057, lines 3 Thus a Trained model reads on a learning model)
Claim 7, Brown in view of Ferrucci discloses the method of claim 1, wherein the query comprises any of a natural language query, an instruction set, a user query, or a system query. (Brown: Section 0009, lines 1-9 understand natural language)
Claim 8, Brown in view of Ferrucci discloses the method of claim 1, further comprising generating a subsequent query based on the respective subsequent output of the multimodal model and respective context; (Brown, Section 0039, lines 14-15- question about the target answer reads on the subsequent prompt) and
wherein the multimodal model determines whether the respective subsequent output of the multimodal model satisfies the query based on the subsequent query. (Brown, Section 0039, lines 21-23- thus uniquely identifying the answer to the query)
Claim 9, Brown in view of Ferrucci discloses the method of claim 1, wherein the respective context comprises any of a concatenation of a plurality of different portions of any of the outputs of the multimodal model. (Brown, Model 1, 2 … Model N shown in Fig. 6 reads on the different portions)
Claim 10, Brown in view of Ferrucci discloses the method of claim 1, further comprising a stopping condition configured to terminate the iterative determinations when the stopping condition is satisfied prior to the query being satisfied. (Brown Section 0010, lines 1-3- thus multiple dialog turns that helps to answer questions).
Claim 11, Brown in view of Ferrucci discloses the method of claim 10, wherein the stopping condition comprises a maximum number of iterations. (Brown: Section 0010, finding the answer after the multiple dialog turns reads on the maximum number of iterations).
Claim 12, Brown in view of Ferrucci discloses the method of claim 1, further comprising pre-processing, using another multimodal model, the query; and
wherein the multimodal model determines whether the query is satisfied based on the pre-processed query. (Brown Section 0010, lines 1-3- thus multiple dialog turns that helps to answer questions).
Claim 13, Brown discloses a system comprising one or more processing devices and one or more memory devices operably coupled to the one or more processing devices (Section 14-15 database stored in a memory storage system)
the one or more memory devices storing executable code effective to cause the one or more processing devices (Controller 630 shown in fig. 8) to:
generate, using a large language model, (Model 1, 2, ---- Model N shown in Fig. 7) a response to a query; (Answer A1, Answer Z1 and Answer Z2 in fig. 7)
iteratively determine, (Section 0007, lines 6-8 “ranging the questions” from broad to narrow shows the iteratively parsing the input) in response to a query whether one or more subsequent outputs of the large language model satisfies the query;
generating a respective rationale to determine a respective output; (Section 0031, lines 1-3 Fig. 1 which is a system that depicts analyzing input or queries and generate answers by ranking the answers are based on a logical architecture)
NB: this means that the answers (respective output) are generated based on logical or rationale architecture- Also see Section 0038.
(Section 0012, lines 4-5 based on the search select or identify candidate answers to the input query) generate a respective context associated with a respective output of the large language model; (Section 0031, lines 7-9 refers to a query (and its context)) and determine, by the large language model based on the respective context, whether the respective subsequent output satisfies the query. (Section 0033, lines 7-9 query response is generated based one or more of multiple modalities including text, audio etc.)
It is not clear how Brown discloses query in context of the multimodal model. Ferrucci discloses model in some common context. (Section 0093, lines 2 “Task related and generally context dependent”)
Therefore it will be obvious to one ordinary skilled in the art before the effective filing date of the claimed invention to include the teaching of analyzing inputs or queries in context. The motivation is that it makes the answers very accurate.
Claim 14, Brown in view of Ferrucci discloses the system of claim 13, wherein the one or more processing devices (Section 14-15 database stored in a memory storage system) are further configured to: obtain at least a portion of one or more data records based on the query; (Section 0038, lines 17-20, Brown: decomposing an input query into it different components such as grammatical , lexical and syntactic and predicate structure and these components reads on the data records) generate an initial prompt based on the at least a portion of the one or more data records; (Brown: Section 0038, lines 7-9- thus a string and an implicit context or inner questions reads on the initial prompt) and
wherein the large language model determines whether the respective subsequent output of the large language model satisfies the query based on the prompt. (Brown, Section 0039, lines 18-20- thus answer (subsequent output) to an inner question reads on the initial prompt as well).
Claim 15, Brown in view of Ferrucci discloses the system of claim 14, wherein the one or more processing devices are further configured to obtaining at least another portion of the one or more data records based on the respective subsequent output of the large language model; (Brown, Section 0038, lines 18-19- thus the Question Classifier block reads on at least another portion of the plurality of data records)
generating a subsequent prompt based on the respective subsequent output of the large language model and the respective context; (Brown, Section 0039, lines 14-15- question about the target answer reads on the subsequent prompt) and
wherein the large language model determines whether the respective subsequent output satisfies the query based on the subsequent prompt. (Brown, Section 0039, lines 21-23- thus uniquely identifying the answer to the query).
Claim 16, Brown in view of Ferrucci discloses the system of claim 14, wherein each portion of the at least a portion of the one or more data records is associated with an embedding value, (Brown, Section 0059, lines 15-17 thus values 0.46, 0.48 reads on the values) and wherein the at least a portion of the one or more data records is obtained based on the embedding value and an embedding index; (Brown: Section 0043, lines 3-7 thus providing an index into the tree) and wherein a retriever module obtains the at least a portion of the one or more data records, and wherein the retriever module provides the at least a portion of the one or more data records to the large language model. (Brown, Section 0060, lines 5-8 search engine which is performing retrieval function)
Brown discloses a non-transitory computer readable medium comprising instructions that when executed cause at least one processor to generate, using a large language model, (Model 1, 2, ---- Model N shown in Fig. 7) a response to a query; (Answer A1, Answer Z1 and Answer Z2 in fig. 7)
iteratively determine, (Section 0007, lines 6-8 “ranging the questions” from broad to narrow shows the iteratively parsing the input) in response to a query whether one or more subsequent outputs of the large language model satisfies the query; (Section 0012, lines 4-5 based on the search select or identify candidate answers to the input query)
generating a respective rationale to determine a respective output; (Section 0031, lines 1-3 Fig. 1 which is a system that depicts analyzing input or queries and generate answers by ranking the answers are based on a logical architecture)
NB: this means that the answers (respective output) are generated based on logical or rationale architecture- Also see Section 0038.
generate a respective context, based on the respective rationale, associated with a respective output of the large language model; (Section 0031, lines 7-9 refers to a query (and its context)) and
determine, by the large language model based on the respective context, whether the respective subsequent output satisfies the query. (Section 0033, lines 7-9 query response is generated based one or more of multiple modalities including text, audio etc.)
It is not clear how Brown discloses query in context of the multimodal model. Ferrucci discloses model in some common context. (Section 0093, lines 2 “Task related and generally context dependent”)
Therefore it will be obvious to one ordinary skilled in the art before the effective filing date of the claimed invention to include the teaching of analyzing inputs or queries in context. The motivation is that it makes the answers very accurate.
Claim 18, Brown in view of discloses the non-transitory computer readable medium of claim 17, further configured to obtain at least a portion of one or more data records based on the query; (Section 0038, lines 17-20, Brown: decomposing an input query into it different components such as grammatical , lexical and syntactic and predicate structure and these components reads on the data records) and
generate an initial prompt based on the at least a portion of the one or more data records; (Section 0038, lines 7-9- thus a string and an implicit context or inner questions reads on the initial prompt)
wherein the large language model determines whether the respective subsequent output of the large language model satisfies the query based on the prompt. (Brown, Section 0039, lines 18-20- thus answer (subsequent output) to an inner question reads on the initial prompt as well).
Cited Art
The prior art made of record and not relied upon is considered pertinent to applicant's disclosure.
Johnston (US 20110161341) teaches mobile search and more specifically to iterative multimodal interfaces for disambiguating search results. A significant source of complexity in authoring these systems is that communication among components is not standardized and often utilizes ad hoc or proprietary protocols. This makes it difficult or impossible to plug-and-play components from different vendors or research sites and limits the ability of authors to rapidly pull components together to prototype multi-modal systems.
Conclusion
Applicant's amendment necessitated the new ground(s) of rejection presented in this Office action. Accordingly, THIS ACTION IS MADE FINAL. See MPEP § 706.07(a). Applicant is reminded of the extension of time policy as set forth in 37 CFR 1.136(a).
A shortened statutory period for reply to this final action is set to expire THREE MONTHS from the mailing date of this action. In the event a first reply is filed within TWO MONTHS of the mailing date of this final action and the advisory action is not mailed until after the end of the THREE-MONTH shortened statutory period, then the shortened statutory period will expire on the date the advisory action is mailed, and any nonprovisional extension fee (37 CFR 1.17(a)) pursuant to 37 CFR 1.136(a) will be calculated from the mailing date of the advisory action. In no event, however, will the statutory period for reply expire later than SIX MONTHS from the mailing date of this final action.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to Akwasi M Sarpong whose telephone number is (571)270-3438. The examiner can normally be reached Mon-Fri. 8:00am-4:00pm.
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.
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.
/AKWASI M SARPONG/ SPE, Art Unit 2681 01/08/2026
Prosecution Insights