DETAILED ACTION
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 1-20 are pending for examination. Claims 1, 10, and 19 are independent.
Priority
Acknowledgement is made of applicant’s claim to U.S. provisional application 63/518,843 filed on 08/10/2023 for domestic benefit under 35 U.S.C. 119 (e).
Information Disclosure Statement
The information disclosure statement (IDS) is submitted on 12/04/2024. The submission is in compliance with the provisions of 37 CFR 1.97. Accordingly, the information disclosure statement is being considered by the examiner.
Claim Objections
Claims 1, 10, and 19 are objected to because of the following informalities:
The abbreviation "LAAs" is stated to represent the term "language model augmented agents" in Claims 1, 10, and 19, but in applicant's specification [0005] "LAAs" is referred to as "LLM-augmented Autonomous Agents".
Appropriate correction is required.
Claim Rejections - 35 USC § 101
35 U.S.C. 101 reads as follows:
Whoever invents or discovers any new and useful process, machine, manufacture, or composition of matter, or any new and useful improvement thereof, may obtain a patent therefor, subject to the conditions and requirements of this title.
Claims 1-20 are rejected under 35 U.S.C. 101 because the claimed invention is directed to an abstract idea without significantly more.
Step 1: The claim is a process, machine, manufacture, or composition of matter.
In the instant application, Claims 1-9 are directed to a process, Claims 10-18 are directed to a machine, and Claims 19-20 are directed to a machine. Thus, each of the claims falls within one of the four statutory categories (i.e. process, machine, manufacture, or composition of matter).
With respect to Claim 1:
2A Prong 1: The claim recites an abstract idea, law of nature, or natural phenomenon.
selecting, by the controller, a LAA from the plurality of LAAs based on the task instruction and the observation; (This step of selecting an agent from a plurality of agents is practically performable in the human mind and is understood to be a recitation of a mental process with the aid of pen and paper (i.e. judgement).)
determining, by the controller, the action based on the output, and (This step of determining an action based on output is practically performable in the human mind and is understood to be a recitation of a mental process with the aid of pen and paper (i.e. evaluation).)
2A Prong 2: The judicial exception is not integrated into a practical application.
A method of predicting an action by a plurality of language model augmented agents (LAAs), the method comprising:
receiving, via a data interface, a task instruction to be performed using an environment; (Receiving information is understood as insignificant extra-solution activity — see MPEP 2106.05(g).)
receiving, by a controller from the environment, an observation of a first state of the environment; (Receiving information is understood as insignificant extra-solution activity — see MPEP 2106.05(g).)
obtaining an output from the selected LAA generated using an input combining the task instruction, the observation, and an LAA-specific prompt template; (Obtaining information is understood as insignificant extra-solution activity — see MPEP 2106.05(g).)
causing the action to be performed on the environment thereby causing the first state of the environment to change to a second state. (Selecting the target of an action is understood to be reciting no more than the idea of a solution or outcome, analogous to the recitation of the words “apply it” (or an equivalent), or as mere instructions to perform the abstract idea (i.e. judgement) — see MPEP 2106.05(f).)
The additional elements as disclosed above alone or in combination do not integrate the judicial exception into practical application as they are insignificant extra-solution activities in combination with mere instructions to apply an exception to perform the abstract idea above.
2B: The claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception.
A method of predicting an action by a plurality of language model augmented agents (LAAs), the method comprising:
receiving, via a data interface, a task instruction to be performed using an environment; (Receiving information is understood as a well-understood, routine, and conventional function, being exemplary of receiving or gathering data — see MPEP 2106.05(d)(II)(i).)
receiving, by a controller from the environment, an observation of a first state of the environment; (Receiving information is understood as a well-understood, routine, and conventional function, being exemplary of receiving or gathering data — see MPEP 2106.05(d)(II)(i).)
obtaining an output from the selected LAA generated using an input combining the task instruction, the observation, and an LAA-specific prompt template; (Obtaining information is understood as a well-understood, routine, and conventional function, being exemplary of receiving or gathering data — see MPEP 2106.05(d)(II)(i).)
causing the action to be performed on the environment thereby causing the first state of the environment to change to a second state. (Selecting the target of an action is understood to be reciting no more than the idea of a solution or outcome, analogous to the recitation of the words “apply it” (or an equivalent), or as mere instructions to perform the abstract idea (i.e. judgement) — see MPEP 2106.05(f).)
The additional elements as disclosed above alone or in combination do not recite significantly more than a judicial exception as they are well-understood, routine, conventional activities previously known to the industry in combination with mere instructions to apply an exception to perform the abstract idea above.
With respect to Claim 10:
see the rejection of Claim 1 above; the same rationale is applied.
2A Prong 2 & 2B: The claim recites another element “A system for predicting an action by a plurality of language model augmented agents (LAAs), the system comprising: a memory that stores the plurality of LAAs and a plurality of processor executable instructions; a communication interface that receives a task instruction to be performed using an environment; and one or more hardware processors that read and execute the plurality of processor-executable instructions from the memory to perform operations comprising:” (The system, memory, communication interface, and one or more processors are understood as mere instructions to apply the exception using a generic computer component — see MPEP 2106.05(f).)
With respect to Claim 19:
see the rejection of Claim 1 above; the same rationale is applied.
2A Prong 2 & 2B: The claim recites another element “A non-transitory machine-readable medium comprising a plurality of machine-executable instructions which, when executed by one or more processors, are adapted to cause the one or more processors to perform operations comprising:” (The non-transitory machine-readable medium is understood as mere instructions to apply the exception using a generic computer component — see MPEP 2106.05(f).)
With respect to Claims 2, 11, and 20:
2A Prong 1: The claim recites an abstract idea, law of nature, or natural phenomenon.
wherein the selecting the LAA from the plurality of LAAs comprises selecting the LAA based on an available action presented by the environment determined by the controller based on the observation of the first state. (This step of selecting an agent from a plurality of agents is practically performable in the human mind and is understood to be a recitation of a mental process with the aid of pen and paper (i.e. judgement).)
2A Prong 2 & 2B: The claim does not recite any additional elements.
With respect to Claims 3 and 12:
2A Prong 1: The claim recites an abstract idea, law of nature, or natural phenomenon.
wherein the selecting the LAA from the plurality of LAAs is (This step of selecting an agent from a plurality of agents is practically performable in the human mind and is understood to be a recitation of a mental process with the aid of pen and paper (i.e. judgement).)
2A Prong 2 & 2B:
wherein the selecting the LAA from the plurality of LAAs is performed by a neural network based model (This step is reciting a judicial exception with the words “apply it” (or an equivalent), or merely invoking computers or machinery as a tool to perform the abstract idea (i.e. evaluation) — see MPEP 2106.05(f).)
With respect to Claims 4 and 13:
2A Prong 1: The claim does not recite an Abstract idea.
2A Prong 2 & 2B:
wherein the output from the selected LAA comprises a recommended action. (The specification of the output is understood to be a field of use limitation — see MPEP 2106.05(h).)
With respect to Claims 5 and 14:
2A Prong 1: The claim does not recite an Abstract idea.
2A Prong 2 & 2B:
wherein the output from the selected LAA comprises information relating to the performance of one or more past actions performed on the environment. (The specification of the output is understood to be a field of use limitation — see MPEP 2106.05(h).)
With respect to Claims 6 and 15:
2A Prong 1: The claim does not recite an Abstract idea.
2A Prong 2 & 2B:
wherein each LAA of the plurality of LAAs is implemented on a neural network based language model. (The specification of the LAA is understood to be a field of use limitation — see MPEP 2106.05(h).)
With respect to Claims 7 and 16:
2A Prong 1: The claim does not recite an Abstract idea.
2A Prong 2 & 2B:
wherein the plurality of LAAs are hosted on one or more external servers. (This step is reciting a judicial exception with the words “apply it” (or an equivalent), or merely invoking computers or machinery as a tool to perform the abstract idea (i.e. evaluation) — see MPEP 2106.05(f).)
With respect to Claims 8 and 17:
2A Prong 1: The claim does not recite an Abstract idea.
2A Prong 2 & 2B:
wherein the plurality of LAAs are hosted on a same server as the controller. (This step is reciting a judicial exception with the words “apply it” (or an equivalent), or merely invoking computers or machinery as a tool to perform the abstract idea (i.e. evaluation) — see MPEP 2106.05(f).)
With respect to Claims 9 and 18:
2A Prong 1: The claim recites an abstract idea, law of nature, or natural phenomenon.
determining, (This step of determining a feedback result is practically performable in the human mind and is understood to be a recitation of a mental process with the aid of pen and paper (i.e. evaluation).)
determining, by the controller together with one or more of the plurality of LAAs, a subsequent action in response to the determination. (This step of determining a subsequent action based on a determination is practically performable in the human mind and is understood to be a recitation of a mental process with the aid of pen and paper (i.e. evaluation).)
2A Prong 2: The judicial exception is not integrated into a practical application.
collecting a feedback from the environment after the action is performed on the environment; (Collecting feedback is understood as insignificant extra-solution activity — see MPEP 2106.05(g).)
determining, by the controller, (This step is reciting a judicial exception with the words “apply it” (or an equivalent), or merely invoking computers or machinery as a tool to perform the abstract idea (i.e. evaluation) — see MPEP 2106.05(f).)
2B: The claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception.
collecting a feedback from the environment after the action is performed on the environment; (This step of collecting feedback after an action is understood as a well-understood, routine, and conventional activity, being exemplary of gathering statistics — see MPEP 2106.05(d)(II)(iv).)
determining, by the controller, (This step is reciting a judicial exception with the words “apply it” (or an equivalent), or merely invoking computers or machinery as a tool to perform the abstract idea (i.e. evaluation) — see MPEP 2106.05(f).)
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, 6-8, 10-11, 15-17, 19-20 are rejected under 35 U.S.C. 103 as being unpatentable over Schillace et al. (US 2024/0202452 A1), hereinafter "Schillace", in view of Das et al. (US 2023/0394413 A1), hereinafter “Das”.
With respect to Claim 1:
Schillace teaches: A method of predicting an action by a plurality of language model augmented agents (LAAs), the method comprising:
receiving, via a data interface, a task instruction to be performed using an environment; ([0023] Schillace discloses “A task objective module receives the task request and determines a task objective of the task request. [0024] discloses “The prompt generator may generate one or more prompts that when processed a ML model, such as a generative large language model (LLM), provide sufficient context for the ML model to generate model output responsive to the task objective associated with the input.”)
receiving, by a controller from the environment, an observation of a first state of the environment; ([0023] Schillace discloses “The semantic encoding model may be utilized to determine the semantic context associated with the intent and task request and determine the task objective… In this way, the ML model gains necessary context to know about the user’s work communications and provide a more compete and relevant output in response to the task request.”)
selecting, by the controller, a LAA from the plurality of LAAs([0027] Schillace discloses “The semantically associated prompt templates are used to generate one or more prompts which will be processed by one or more ML models from the model repository. In further embodiments, a ML model stored in model repository may be trained to output prompts. The trained ML model may be utilized to process the intent and task objective and output one or more prompts responsive to the input.”)
obtaining an output from the selected LAA generated using an input combining the task instruction, the observation, and an LAA-specific prompt template; ([0024] Schillace discloses “the one or more prompts encompass the semantic context of the task objective and task request so that the ML model can generate model output responsive to the requested task and/or intent without requiring additional training or fine-tuning of the model prior to generating model output responsive to the task or intent. It will be appreciated that a prompt may be comprised of a plurality of prompt templates. A prompt template may include any of a variety of data… the type of data may depend on the type of ML model that will be leveraged to respond to the received input.” )
determining, by the controller, the action based on the output, and ([0022] Schillace discloses “the request processor may provide the model output to the application after it is generated and/or evaluated by the response evaluator.” [0029] discloses “In some aspects, if the model output is determined to be unresponsive to the input, the response evaluator may reinitiate the process for generating the model output,”)
causing the action to be performed on the environment thereby causing the first state of the environment to change to a second state. ([0060] Schillace discloses “model output may be associated with a corresponding application and/or data format, such that model output is processed to display the output to a user and/or to fabricate a physical object, among other examples.”)
Schillace does not teach:
selecting, by the controller, a LAA from the plurality of LAAs based on the task instruction and the observation;
However, Das teaches in the same field of endeavor:
selecting, by the controller, a LAA from the plurality of LAAs based on the task instruction and the observation; ([0024] Das discloses “the reward maximization module may select optimal agents for a particular task based on contextual information about agents, environment and intent (e.g., a problem that needs to be solved by a team).”)
Schillace and Das are both analogous art to the present invention because both are from the same field of endeavor directed towards prompting the selection of specific language models to generate an output directed towards taking relevant action regarding the target task.
It would have been obvious for one of ordinary skill in the art prior to the effective filing date of the claimed invention to modify Schillace’s teachings by using the prompt and environment or contextual information to support the selection of a task-specific agent to generate the target output or perform the target task as taught by Das. One would have been motivated to make this modification in order to improve the output of the agent and to make that output more specific to the task at hand.
With respect to Claim 10:
Schillace in view of Das teaches: A system for predicting an action by a plurality of language model augmented agents (LAAs), the system comprising:
a memory that stores the plurality of LAAs and a plurality of processor executable instructions; ([0017, 0033] Schillace discloses the response engine that encompasses a request wrapper and model repository for storing LAAs, the response engine having access to the data store that may be a type of memory.)
a communication interface that receives a task instruction to be performed using an environment; and ([0019] Schillace discloses “the input may be received, for example, in a chat function of application(s) used for interacting with a ML model in model repository,” wherein the char function is understood as a communication interface where task instructions may be received.)
one or more hardware processors that read and execute the plurality of processor-executable instructions from the memory to perform operations comprising: ([0032] Schillace discloses at least one processor that executes software and/or firmware stored in memory, the software/firmware code containing instructions that can be executed by the processor.)
Claim 10 is a device claim that corresponds to Claim 1 and the remaining limitations are rejected for at least the same reasons therein.
With respect to Claim 19:
Schillace in view of Das teaches: A non-transitory machine-readable medium comprising a plurality of machine-executable instructions which, when executed by one or more processors, are adapted to cause the one or more processors to perform operations comprising: ([0036] Schillace discloses “The method can be executed as computer-executable instructions executed by a computer system and encoded or stored on a computer readable medium or other non-transitory computer storage media.)
Claim 19 is a non-transitory machine-readable medium claim that corresponds to Claim 1 and the remaining limitations are rejected for at least the same reasons therein.
With respect to Claims 2, 11, and 20:
Schillace in view of Das teaches: The method of Claim 1, system of Claim 10, and non-transitory machine-readable medium of Claim 19 respectively, wherein the selecting the LAA from the plurality of LAAs comprises selecting the LAA based on an available action presented by the environment determined by the controller based on the observation of the first state. ([0024] Das discloses “the reward maximization module may select optimal agents for a particular task based on contextual information about agents, environment and intent.”)
With respect to Claims 6 and 15:
Schillace in view of Das teaches: The method of Claim 1 and system of Claim 10 respectively, wherein each LAA of the plurality of LAAs is implemented on a neural network based language model. ([0028] Schillace discloses that the model repository may include any of a variety of ML models, including generative models such as large language models, which are built on neural network architecture.)
With respect to Claims 7 and 16:
Schillace in view of Das teaches: The method of Claim 1 and system of Claim 10 respectively, wherein the plurality of LAAs are hosted on one or more external servers. ([0022 – 0024] Das discloses how the models in the machine learning system are executed by the computing system which may be implemented as “any suitable computing system, such as one or more server computers… in some examples, computing system may represent a cloud computing system, server farm, and/or server cluster,”)
With respect to Claims 8 and 17:
Schillace in view of Das teaches: The method of Claim 1 and system of Claim 10 respectively, wherein the plurality of LAAs are hosted on a same server as the controller. ([0022-0024] Das discloses how the models and multi-agent controllers are both generated by the machine learning system and executed by the computing system which may be implemented as “any suitable computing system, such as one or more server computers,”)
Claims 3-5, 9, 12-14, and 18 are rejected under 35 U.S.C. 103 as being unpatentable over Schillace in view of Das and Liguori et al. (US 12,524,214 B1), hereinafter “Liguori”.
With respect to Claims 3 and 12:
Schillace in view of Das teaches: The method of Claim 1 and system of Claim 10 respectively, Schillace in view of Das does not teach:
wherein the selecting the LAA from the plurality of LAAs is performed by a neural network based model predicting which one of the plurality of LAAs is to be employed based on an input of the task instruction and the observation of the first state.
However, Liguori teaches in the same field of endeavor:
wherein the selecting the LAA from the plurality of LAAs is performed by a neural network based model predicting which one of the plurality of LAAs is to be employed based on an input of the task instruction and the observation of the first state. ([Fig. 8, Col. 12 Lines 37-43] Liguori discloses how the orchestrator agent identifies which agent is best to deploy for the task completion by using the initial prompt that encompasses a task instruction. [Col. 18 Lines 47-48, Col. 6 Lines 14-20] Liguori discloses how those prompts received by the orchestrator agent may be based on prompt templates that include context-specific information gathered from a user or from other resources available to the agent. [Col. 11 Lines 50-55] Liguori discloses “These models (LLMs) are trained using machine learning techniques, typically on vast amounts of text data from the internet, books, articles, and other sources… LLMs use a type of neural network called a transformer to process and understand the patterns and structures of language.”)
Schillace, Das, and Liguori are all analogous art to the present invention because they are from the same field of endeavor directed towards leveraging a language model to generate an output directed towards taking relevant action regarding the target task.
It would have been obvious for one of ordinary skill in the art prior to the effective filing date of the claimed invention to modify Schillace in view of Das’s teachings by using a neural network model to guide the selection of an agent that outputs a recommended action to address the task as taught by Liguori. One would have been motivated to make this modification in order to ensure greater success of the agent actions through utilizing a feedback loop.
With respect to Claims 4 and 13:
Schillace in view of Das and Liguori teaches: The method of Claim 1 and system of Claim 10 respectively, wherein the output from the selected LAA comprises a recommended action. ([Col. 25 Lines 37-39] Liguori discloses “the development agent can send a prompt including an action or step from the action plan to a code generation model,” [Col. 25 Lines 53-57] Liguori discloses “the development task agent sends the action plan to the client. The development agent can include within the action plan, for each action, the recommended code change(s), if any, obtained from the code generation model.”)
With respect to Claims 5 and 14:
Schillace in view of Das and Liguori teaches: The method of Claim 1 and system of Claim 10 respectively, wherein the output from the selected LAA comprises information relating to the performance of one or more past actions performed on the environment. ([Col. 13 Lines 20-27] Liguori discloses using context aggregators to gather additional context after receiving previous prompts from the LLM and sending the updated information in subsequent prompts to the LLM, suggesting that the agent can update prompts based on new environmental context after already prompting the LLM to take an initial action. [Col. 26 Lines 17-32] Liguori discloses receiving changes to a software system and updating the software system based on obtained data and using an action plan from the LLM comprising one or more steps to implement the changes, wherein the changes may occur upon review of a past action from the agent.)
With respect to Claims 9 and 18:
Schillace in view of Das and Liguori teaches: The method of Claim 1 and system of Claim 10 respectively, further comprising:
collecting a feedback from the environment after the action is performed on the environment; ([Col. 6 Lines 1-12] Liguori discloses error logging and an error resolution agent that gathers details regarding the error, which may entail details of the environment. [Col. 17 Lines 1-12] disclose a validation agent which prompts to see if the action performed complies with the requirements.)
determining, by the controller, that the feedback indicates the action was unsuccessful to achieve a desired goal corresponding to the task instruction; and ([Col. 17 Lines 15-19, 22-25] Liguori discloses the validation agent receiving a negative response indicating that the action was unsuccessful to achieve the desired goal. The validation agent may also receive an explanation of why the response or action was unsuccessful. [Col. 17 Lines 37-47] Liguori discloses similar teachings that apply to execution-based validations as opposed to LLM-based validations.)
determining, by the controller together with one or more of the plurality of LAAs, a subsequent action in response to the determination. ([Col. 17 Lines 15-19, 30-33] Liguori discloses the validation agent sending a response correction prompt to the LLM, which may include a recommended subsequent action to take in response to the received negative response from the LLM, and this process may be looped continuously until the response or action achieves the desired goal. [Col. 17 Lines 37-47] Liguori discloses similar teachings that apply to execution-based validations as opposed to LLM-based validations.)
Conclusion
Any inquiry concerning this communication or earlier communications from the examiner should be directed to BRIAN D. BUI whose telephone number is (571)270-0463. The examiner can normally be reached Monday - Friday 8:00am - 5:00pm.
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.
If attempts to reach the examiner by telephone are unsuccessful, the examiner’s supervisor, ABDULLAH AL KAWSAR can be reached at (571) 270-3169. 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.
/BRIAN D. BUI/Examiner, Art Unit 2127 /ABDULLAH AL KAWSAR/Supervisory Patent Examiner, Art Unit 2127
Prosecution Insights