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 .
Claims 1 – 13 and 25 – 37 are pending.
Election/Restrictions
Applicant’s election without traverse of Invention I in the reply filed on 22 June 2026 is acknowledged.
Claim Rejections - 35 USC § 102
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 the appropriate paragraphs of 35 U.S.C. 102 that form the basis for the rejections under this section made in this Office action:
A person shall be entitled to a patent unless –
(a)(2) the claimed invention was described in a patent issued under section 151, or in an application for patent published or deemed published under section 122(b), in which the patent or application, as the case may be, names another inventor and was effectively filed before the effective filing date of the claimed invention.
Claim(s) 1 – 13 and 25 - 37 is/are rejected under 35 U.S.C. 102(a)(2) as being anticipated by US 2023/0153546 A1 issued to Peleg et al (hereinafter Peleg).
As to claim 1, Peleg discloses a system comprising:
a processor (processing devices, see Peleg: Para. 0008 – 0019);
a large language model (trained neural network-based language models for the writing assistant models, see Peleg: Para. 0179 – 0204) configured to generate a response to a prompt context (writing assistant used for providing response to user input prompts, see Peleg: Para. 0112 – 0120 and 0179 – 0204), the large language model including a query generating portion and a response generating portion (using user prompt to search/query external sources using search engines, social medial and/or other data mining techniques to generate responses, see Peleg: Para. 0112 – 0120 and 0179 – 0204), the query generating portion receiving the prompt context as input and providing a generated query as output (using user input prompt, preexisting text context when querying the external sources via search engine, social media and other data mining techniques, see Peleg: Para. 0112 – 0120 and 0179 – 0204) and the response generating portion receiving the prompt context and a plurality of encoded context passages as input and providing a generated response as output (generating text output using the text input, text and contextual clues from the identified documents, including verified facts from the documents of the external knowledge bases, see Peleg: Para. 0112 – 0120 and 0179 – 0204, and encoding position and length of text tokens in the source text, and using the encoded position and length of tokens to generate conditioned output tokens in the response, see Peleg: Para. 0170), wherein the response generating portion is trained to determine whether or not to use the plurality of encoded context passages (determining whether to replace the text output with corrections of verified facts from the external knowledge base, see Peleg: Para. 0112 – 0120 and 0179 – 0204, and encoding position and length of text tokens in the source text, and using the encoded position and length of tokens to generate conditioned output tokens in the response, see Peleg: Para. 0170); and
memory storing instructions that, when executed by the processor, cause the system to:
receive a first prompt context, the first prompt context including a prompt from a user (receive first user prompt, such as “Tony Johnson” or “Bono’s age is?” or “propose meeting” or “introduce someone”, see Peleg: Para. 0112 - 0120),
receive a first generated query by providing the first prompt context to the query generating portion (the writing assistant can send the prompt to search engines, social media and/or use other type of data mining to identify documents from external sources, see Peleg: Para. 0112 – 0120),
receive a set of supporting resources by providing the first generated query to a search engine, the set of supporting resources being documents identified as responsive to the first generated query (the writing assistant can send the prompt to search engines, social media and/or use other type of data mining to identify documents from external sources, see Peleg: Para. 0112 – 0120),
generate a first plurality of encoded context passages based on the set of supporting resources (relevant text from the identified documents from the external sources is identified using the input prompt and contextual aspects of the document, see Peleg: Para. 0112 – 0120, and encoding position and length of text tokens in the source text, and using the encoded position and length of tokens to generate conditioned output tokens in the response, see Peleg: Para. 0170),
generate a first response to the first prompt context by providing the first plurality of encoded context passages and the first prompt context to the response generating portion (text output options are generated by the writing assistant, including providing answers to user prompts, additional data and facts related to the input, or corrections of factual errors included in the user prompt or that exist in preexisting text, and using contextual clues from the documents and external knowledge base, see Peleg: Para. 0112 – 0120, and encoding position and length of text tokens in the source text, and using the encoded position and length of tokens to generate conditioned output tokens in the response, see Peleg: Para. 0170), and
provide the first response for presentation to the user (provide the text output to the user, see Peleg: Para. 0112 – 0120).
As to claim 2, Peleg discloses the system as in claim 1, wherein the plurality of encoded context passages include, for each supporting resource, text selected by the search engine from content of the supporting resource (identifying text in the external sources for use in the generation of output, also verification of generated output text using external knowledge bases, see Peleg: Para. 0112 – 0120, and encoding position and length of text tokens in the source text, and using the encoded position and length of tokens to generate conditioned output tokens in the response, see Peleg: Para. 0170).
As to claim 3, Peleg discloses the system as in claim 1, the memory further including instructions that cause the system to:
for each supporting resource in the set of supporting resources:
identify at least two relevant portions of content (identifying and selecting portions of text from the documents of the sources, see Peleg: Para. 0372 - 0373), and
concatenate the at least two relevant portions into a context passage for the supporting resource (organize selected portions of text into logical order that flow together in a coherent and fluent manner, see Peleg: Para. 0373),
wherein the first encoded context passages are generated from the context passages for the set of supporting resources (identifying text in the external sources for use in the generation of output, also verification of generated output text using external knowledge bases, see Peleg: Para. 0112 – 0120, and encoding position and length of text tokens in the source text, and using the encoded position and length of tokens to generate conditioned output tokens in the response, see Peleg: Para. 0170).
As to claim 4, Peleg discloses the system as in claim 3, wherein a relevant portion of the at least two relevant portions has a length that is longer than text selected by the search engine from the content of the supporting resource (organize selected portions of text into logical order that flow together in a coherent and fluent manner, including adding additional words or text, developing linking phrases, transitional phrases, clauses or sentences, see Peleg: Para. 0373, organization of multiple selected portions is longer than each individual portion, and the additional words/phrases/etc. make the output longer than the combination of original portions of text).
As to claim 5, Peleg discloses the system as in claim 4, wherein the length is less than 2000 characters (example responses are less than 2000 characters, see Peleg: Para. 0052, 0066, 0083, 0112 – 0120, 0394 – 0437, and 10 word limit, see Para. 0097).
As to claim 6, Peleg discloses the system as in claim 1, the memory further including instructions that cause the system to:
for each supporting resource in the set of supporting resources, identify a relevant portion of content of the supporting resource, the relevant portion having a length longer than text selected by the search engine for the supporting resource (identifying and selecting portions of text from the documents of the sources, organize selected portions of text into logical order that flow together in a coherent and fluent manner as well as included full documents, see Peleg: Para. 0372 – 0374),
wherein the first encoded context passages are generated from the relevant portions for the set of supporting resources (organize selected portions of text into logical order that flow together in a coherent and fluent manner, including adding additional words or text, developing linking phrases, transitional phrases, clauses or sentences, see Peleg: Para. 0373).
As to claim 7, Peleg discloses the system as in claim 6, the memory further including instructions that cause the system to, for at least one supporting resource of the set of supporting resources:
identify two relevant portions of content of the supporting resource, each relevant portion having a length longer than text selected by the search engine for the supporting resource (identifying and selecting portions of text from the documents of the sources, organize selected portions of text into logical order that flow together in a coherent and fluent manner as well as included full documents, see Peleg: Para. 0372 – 0374); and
concatenate the two relevant portions to form a context passage, the context passage being used to generate the encoded context passages (organize selected portions of text into logical order that flow together in a coherent and fluent manner, including adding additional words or text, developing linking phrases, transitional phrases, clauses or sentences, see Peleg: Para. 0373).
As to claim 8, Peleg discloses the system as in claim 1, the memory further including instructions that cause the system to:
determine that the prompt relates to a fact (identifying facts from generated output and verifying the facts using external knowledge bases, and correcting errors in fact in the user input and/or preexisting text, see Peleg: Para. 0112 – 0120); and
provide the first generated query with the first response for presentation to the user (provide the text output to the user, see Peleg: Para. 0112 – 0120).
As to claim 9, Peleg discloses the system as in claim1, the memory further including instructions that cause the system to:
determine that a span of text in the generated response includes a fact (identifying facts from generated output and verifying the facts using external knowledge bases, and correcting errors in fact in the user input and/or preexisting text, see Peleg: Para. 0112 – 0120);
identify, from the set of supporting resources, content in a supporting resource that supports the fact (identifying facts from generated output and verifying the facts using external knowledge bases, and correcting errors in fact in the user input and/or preexisting text, see Peleg: Para. 0112 – 0120);
alter an appearance of the span (correcting errors in the facts after verification of the facts using the external knowledge bases, see Pele: Para. 0112 – 0120); and
provide a control that, when selected, navigates to the supporting resource (adding hyperlinks to each source in the response to the user so that the user may navigate to the external source, see Peleg: Para. 0365).
As to claim 10, Peleg discloses the system as in claim 9, wherein the control is one of a favicon for the supporting resource or a hyperlink having the span as anchor text (adding hyperlinks to each source in the response to the user so that the user may navigate to the external source, see Peleg: Para. 0365, and the source identifier in the response to the user may be hyperlinked, see Peleg: Para. 0370).
As to claim 11, Peleg discloses the system as in claim 1, the memory further including instructions that cause the system to:
determine that a span of text in the generated response includes a fact (identifying facts from generated output and verifying the facts using external knowledge bases, and correcting errors in fact in the user input and/or preexisting text, see Peleg: Para. 0112 – 0120);
identify, from the set of supporting resources, content in a supporting resource that supports the fact (adding hyperlinks to each source in the response to the user so that the user may navigate to the external source, see Peleg: Para. 0365); and
provide a control that, when selected, navigates to the supporting resource (adding hyperlinks to each source in the response to the user so that the user may navigate to the external source, see Peleg: Para. 0365).
As to claim 12, Peleg discloses the system as in claim 1, the memory further including instructions that cause the system to:
determine that a span of text in the generated response includes a fact (identifying facts from generated output and verifying the facts using external knowledge bases, and correcting errors in fact in the user input and/or preexisting text, see Peleg: Para. 0112 – 0120); and
add a hyperlink to the span, the hyperlink configured to submit the span of text as a query to the search engine (adding hyperlinks to each source in the response to the user so that the user may navigate to the external source, see Peleg: Para. 0365).
As to claim 13, Peleg discloses the system as in claim 1, the memory further including instructions that cause the system to:
determine that a span of text in the generated response includes a fact (identifying facts from generated output and verifying the facts using external knowledge bases, and correcting errors in fact in the user input and/or preexisting text, see Peleg: Para. 0112 – 0120);
submit the span of text as a query to the search engine (verify the fact(s) using queries to external knowledge bases, see Peleg: Para. 0112 - 0120);
alter an appearance of the span (correcting errors in the facts after verification of the facts using the external knowledge bases, see Pele: Para. 0112 – 0120); and
provide a control that, when selected, navigates to a top-ranked search result identified by the search engine as responsive to the query (acquired response examples/phrases are displayed to the user based on rank, and adding hyperlinks to each source in the response to the user so that the user may navigate to the external source, see Peleg: Para. 0363 – 0365).
As to claim 25, Peleg discloses a method comprising:
receiving a prompt context, the prompt context including a prompt from a user (receiving user input and determining context of user input, see Peleg: Para. 0112 – 0120),
obtaining a first query from a query generating portion of a large language model, the query generating portion using the prompt context as input (receive first user prompt, such as “Tony Johnson” or “Bono’s age is?” or “propose meeting” or “introduce someone”, the writing assistant can send the prompt to search engines, social media and/or use other type of data mining to identify documents from external sources, see Peleg: Para. 0112 – 0120);
receiving a set of supporting resources by providing the first query to a search engine, the set of supporting resources being documents identified as responsive to the first query (the writing assistant can send the prompt to search engines, social media and/or use other type of data mining to identify documents from external sources, see Peleg: Para. 0112 – 0120);
generating a response to the prompt context by providing relevant content from the set of supporting resources and the prompt context to a response generating portion of the large language model (text output options are generated by the writing assistant, including providing answers to user prompts, additional data and facts related to the input, or corrections of factual errors included in the user prompt or that exist in preexisting text, and using contextual clues from the documents and external knowledge base, see Peleg: Para. 0112 – 0120);
determining that a span of text in the generated response includes a fact (identifying facts from generated output and verifying the facts using external knowledge bases, and correcting errors in fact in the user input and/or preexisting text, see Peleg: Para. 0112 – 0120);
submitting the span of text as a second query to the search engine (searching the identified facts in the generated text out in one or more external knowledge bases to verify the facts of the generated response, see Peleg: Para. 0112 – 0120); and
providing the response (provide the text output to the user, see Peleg: Para. 0112 – 0120) and a control that, when selected, navigates to a top-ranked search result identified by the search engine as responsive to the second query for presentation to the user (acquired response examples/phrases are displayed to the user based on rank, and adding hyperlinks to each source in the response to the user so that the user may navigate to the external source, see Peleg: Para. 0363 – 0365).
As to claim 26, Peleg discloses a method comprising:
receiving a prompt context, the prompt context including a prompt from a user (receiving user input and determining context of user input, see Peleg: Para. 0112 – 0120),
obtaining a query from a query generating portion of a large language model, the query generating portion using the prompt context as input (receive first user prompt, such as “Tony Johnson” or “Bono’s age is?” or “propose meeting” or “introduce someone”, the writing assistant can send the prompt to search engines, social media and/or use other type of data mining to identify documents from external sources, see Peleg: Para. 0112 – 0120);
receiving a set of supporting resources by providing the query to a search engine, the set of supporting resources being documents identified as responsive to the query (the writing assistant can send the prompt to search engines, social media and/or use other type of data mining to identify documents from external sources, see Peleg: Para. 0112 – 0120);
generating a response to tile prompt context by providing relevant content from tile set of supporting resources and the prompt context to a response generating portion of the large language model (text output options are generated by the writing assistant, including providing answers to user prompts, additional data and facts related to the input, or corrections of factual errors included in the user prompt or that exist in preexisting text, and using contextual clues from the documents and external knowledge base, see Peleg: Para. 0112 – 0120);
determining that a span of text in the generated response includes a fact (identifying facts from generated output and verifying the facts using external knowledge bases, and correcting errors in fact in the user input and/or preexisting text, see Peleg: Para. 0112 – 0120);
adding a hyperlink to the span, the hyperlink configured to submit the span of text as a query to the search engine (searching the identified facts in the generated text out in one or more external knowledge bases to verify the facts of the generated response, see Peleg: Para. 0112 – 0120, and adding hyperlinks to each source to the user so that the user may navigate to the external source, see Peleg: Para. 0365); and
providing the response for presentation to the user (provide the text output to the user, see Peleg: Para. 0112 – 0120).
As to claim 27, Peleg discloses the method as in claim 26, wherein, in response to receipt of the span of text as a query, the search engine provides a search result page for the span that is displayed in a panel adjacent to the generated response (text portion options and text output options are displayed in separate panels of the template for selection by the user, see Peleg: Para. 0121 – 0131 and Fig. 6A – 6O).
As to claim 28, Peleg discloses the method as in claim 26, further comprising:
submitting the span of text to the query generating portion to generate a suggested query (identifying facts from generated output, searching the identified facts in the generated text out in one or more external knowledge bases to verify the facts of the generated response, and correcting errors in fact in the user input and/or preexisting text, see Peleg: Para. 0112 – 0120), and
providing the suggested query as a selectable user control displayed with the generated response (provide the text output to the user, see Peleg: Para. 0112 – 0120).
As to claim 29, Peleg discloses the method as in claim 26, wherein generating the response to the prompt context includes:
for each supporting resource in the set of supporting resources:
identifying at least two relevant portions of content (identifying and selecting portions of text from the documents of the sources, see Peleg: Para. 0372 - 0373), and
concatenating the at least two relevant portions into a context passage for the supporting resource (organize selected portions of text into logical order that flow together in a coherent and fluent manner, see Peleg: Para. 0373),
wherein the relevant content includes the context passages for the set of supporting resources (identifying text in the external sources for use in the generation of output, also verification of generated output text using external knowledge bases, see Peleg: Para. 0112 – 0120, and encoding position and length of text tokens in the source text, and using the encoded position and length of tokens to generate conditioned output tokens in the response, see Peleg: Para. 0170).
As to claim 30, Peleg discloses the method as in claim 25, wherein the control includes one of a favicon for the supporting resource (selecting icons for the initiating text review by the writing assistant, see Peleg: Para. 0103, 0160, 0364, 0377, 0399 – 0400, 0452, and source file can be accessed using the provided file icon, see Peleg: Para. 0387).
As to claim 31, Peleg discloses the method as in claim 25, wherein the control includes a hyperlink having the span as anchor text (the source identifier in the response to the user may be hyperlinked, see Peleg: Para. 0370).
As to claim 32, Peleg discloses the method as in claim 25, further comprising:
submitting the span of text to the query generating portion to generate a suggested query (identifying facts from generated output, searching the identified facts in the generated text out in one or more external knowledge bases to verify the facts of the generated response, and correcting errors in fact in the user input and/or preexisting text, see Peleg: Para. 0112 – 0120),
providing the suggested query as a selectable user control displayed with the generated response (provide the text output to the user, see Peleg: Para. 0112 – 0120).
As to claim 33, Peleg discloses the method as in claim 25, further comprising:
determining, prior to obtaining the first query, that the prompt includes a reference to an entity (determing the user prompt is for a person, such as “Tony Johnson”, and identifying data about the person such as gender and other facts associated with the individual, see Peleg: Para. 0112 – 0120);
identifying the entity based on metadata or on a prior prompt within the prompt context (identifying data about the person such as gender and other facts associated with the individual, see Peleg: Para. 0112 – 0120, and identifying data based on metadata fields and prefix tokens from prior system samples, see Peleg: Para. 0193); and
updating the prompt to include the entity (using the user input prompt with the contextual aspects to the search the external sources for additional data/facts about the person, see Peleg: Para. 0112 – 0120).
As to claim 34, Peleg discloses a method comprising:
receiving a first prompt context, the first prompt context including a prompt from a user (receiving user input and determining context of user input, see Peleg: Para. 0112 – 0120),
receiving a first generated query by providing the first prompt context to a query generating portion of a large language model (receive first user prompt, such as “Tony Johnson” or “Bono’s age is?” or “propose meeting” or “introduce someone”, the writing assistant can send the prompt to search engines, social media and/or use other type of data mining to identify documents from external sources, see Peleg: Para. 0112 – 0120, and trained neural network-based language models for the writing assistant models, see Peleg: Para. 0179 – 0204),
receiving a set of supporting resources for the first generated query, the set of supporting resources being documents identified as responsive to the first generated query (the writing assistant can send the prompt to search engines, social media and/or use other type of data mining to identify documents from external sources, see Peleg: Para. 0112 – 0120),
generating a first plurality of context passages based on the set of supporting resources (relevant text from the identified documents from the external sources is identified using the input prompt and contextual aspects of the document, see Peleg: Para. 0112 – 0120),
generating a first response to the first prompt context by providing the first plurality of context passages and the first prompt context to a response generating portion of the large language model (text output options are generated by the writing assistant, including providing answers to user prompts, additional data and facts related to the input, or corrections of factual errors included in the user prompt or that exist in preexisting text, and using contextual clues from the documents and external knowledge base, see Peleg: Para. 0112 – 0120), and
providing the first response for presentation to the user (provide the text out to the user, see Peleg: Para. 0112 – 0120).
As to claim 35, Peleg discloses the method as in claim 34, wherein the response generating portion is trained to determine whether or not to use the plurality of context passages (determining whether to replace the text output with corrections of verified facts from the external knowledge base, see Peleg: Para. 0112 – 0120 and 0179 – 0204, and encoding position and length of text tokens in the source text, and using the encoded position and length of tokens to generate conditioned output tokens in the response, see Peleg: Para. 0170).
As to claim 36, Peleg discloses the method as in claim 34, further comprising:
for each supporting resource in the set of supporting resources:
identifying at least two relevant portions of content (identifying and selecting portions of text from the documents of the sources, see Peleg: Para. 0372 - 0373), and
concatenating the at least two relevant portions into a context passage for the supporting resource (organize selected portions of text into logical order that flow together in a coherent and fluent manner, see Peleg: Para. 0373),
wherein the first context passages are generated from the context passages for the set of supporting resources (identifying text in the external sources for use in the generation of output, also verification of generated output text using external knowledge bases, see Peleg: Para. 0112 – 0120, and encoding position and length of text tokens in the source text, and using the encoded position and length of tokens to generate conditioned output tokens in the response, see Peleg: Para. 0170).
As to claim 37, Peleg discloses the method as in claim 34, further comprising:
determining that a span of text in the generated response includes a fact (identifying facts from generated output and verifying the facts using external knowledge bases, and correcting errors in fact in the user input and/or preexisting text, see Peleg: Para. 0112 – 0120);
identifying, from the set of supporting resources, content in a supporting resource that supports the fact (verifying the facts of the generated output using the one or more external knowledge bases, and correcting any errors in fact, see Peleg: Para. 0112 – 0120); and
providing a control that, when selected, navigates to the supporting resource (adding hyperlinks to each source in the response to the user so that the user may navigate to the external source, see Peleg: Para. 0365).
Conclusion
Any inquiry concerning this communication or earlier communications from the examiner should be directed to MARK E HERSHLEY whose telephone number is (571)270-7774. The examiner can normally be reached M-F: 9am-6pm.
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, Amy Ng can be reached at (571) 270-1698. 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.
/MARK E HERSHLEY/Primary Examiner, Art Unit 2164
Prosecution Insights