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 .
This action is in response to the communication filed on February 19, 2025.
Claims 1-13 are pending in this action.
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 1-5 and 12-13 are rejected under 35 U.S.C. 101 because
the claimed invention is directed to a judicial exception (i.e., law of nature, a natural
phenomenon, or an abstract idea) without significantly more. The claim(s) recite(s) an
abstract idea of training a generative language model. The claim(s) does/do not
include additional elements that are sufficient to amount to significantly more than the
judicial exception because the claims directed to an abstract idea of training a generative language model. The claim is drawn to process (a series of steps or acts, i.e., constructing learning data having a triple form …, training a generative language model …, and output a next word string of input context corresponding to detected input context vector) that similar to an idea ‘Of itself such as an instantiated concept, plan or scheme, as well as a mental process (thinking) that “can be performed in the human mind, or by a human using a pen and paper’.
The claim does not require that the method be implemented by a particular machine. The method does not require a particular transformation of a particular article. There is not transformation of a physical objects or data into a different state or thing. This training a generative language model is similar to delivering user-selected media content to a portable devices found by the courts to be abstract idea (Affinity Labs of Tex., LLC v. Amazon.com Inc., 120 USPQ2d 1210 (Fed. Cir. 2016)) and also displaying certain results of the collection and analysis found by the courts to be abstract idea (Elec. Power Grp., LLC v. Alstom S.A., 119 USPQ2d 1739 (Fed. Cir. 2016).
This judicial exception is not integrated into a practical application because claims broadly recite the result (training a generative language model, constructing learning data having a triple form …, training a generative language model …, and output a next word string of input context corresponding to detected input context vector), rather than sufficiently claiming a technical means of achieving the result. See Two-Way Media Ltd. v. Comcast Cable Commons, LLC, 874 F.3d 1329, 1337 (Fed. Cir. 2017) (“The claim requires the functional results ... but does not sufficiently describe how to achieve these results in a non-abstract way.”).
The claims recite a Judicial exception relating to “training a generative language model, constructing learning data having a triple form …, training a generative language model …, and output a next word string of input context corresponding to detected input context vector”. Here the claims do not change the underlying or other technology, rather the claimed techniques playing using computing device as pedagogical tool. The claimed additional elements - -computer- -“merely use a computer as a tool to perform an abstract idea” or “do no more than generally link the use of a judicial exception to a particular technological environment.” Memorandum, 84 Fed. Reg. at 55; see Customedia Techs., LLC v. Dish Network Corp., No. 2018- 2239, 2020 WL 1069742, at *3 (Fed. Cir. Mar. 6, 2020) (“We have held that it is not enough, however, to merely improve a fundamental practice or abstract process by invoking a computer merely as a tool.”).
Accordingly, claims 1-5 and 12-13 do not integrate the judicial exception into a practical application. See Memorandum, 84 Fed. Reg. at 54. As the claim recites a judicial exception and fails to integrate the exception into a practical application, the claim is “directed to the .. . judicial exception.” Id. at 54.
The claim(s) does/do not include additional elements that are sufficient to amount to significantly more than the judicial exception because the additional elements are simply a generic computer. The claim amounts to no more than training a generative language model, constructing learning data having a triple form …, training a generative language model …, and output a next word string of input context corresponding to detected input context vector. Taking the claimed elements either individually or as ordered combination, that transform claims into patent-eligible application, since claims merely recite use of already existing computer-based training a generative language model, and there is no “inventive concept” in play using computing device well- understood, routine, and conventional activities commonly used in industry training generative language model, since claims, at most, attempt to limit abstract idea to particular technological environment, and such limitation has been held insufficient to save claims in this context, and since dependent claims are not rendered patent-eligible by recitation of additional steps, such as per claim 2, recites, “separating learning data having a sentence from in an arbitrary chunk …”; as per claim 3, recites, “constructing a vocabulary set having a minimum word unit . . .”; even though additional limitations may narrow scope of claims. The claim as a whole does not amount to significantly more than the abstract idea itself. Accordingly, claims 1-5 and 12-13, are ineligible.
Claim Rejections - 35 USC § 102
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, 12, and 13 is/are rejected under 35 U.S.C. 102(a)(2) as being anticipated by Liu et al. (US 2024/0320447).
As per claims 1 and 12, Liu discloses, a method of training a generative language model, which is performed by a computer, the method comprising:
constructing learning data having a triple form and comprising input context, a vector of the input context (hereinafter referred to as an "input context vector"), and a next word string of the input context (Fig. 2, element 210 “input context”, element 225 “context vector”, and element 235 “next word”); and
training a generative language model to convert a previous output sentence into an input context vector, detect an input context vector that is most similar to the converted input context vector in the learning data having the triple form (Paragraphs 0053-0058), and
output a next word string of input context corresponding to the detected input context vector (Paragraphs 0059-0060).
As per claim 13, Liu discloses, an inference method of a generative language model, which is performed by a computer, the inference method comprising:
generating, as a query context vector, input context received from a user by encoding the input context (Paragraph 0041);
detecting K support context vectors closest to the query context vector in learning data having a triple form and comprising input context, a vector of the input context (hereinafter referred to as an "input context vector"), and a next word string of the input context (Paragraphs 0041-0045);
connecting a next word string corresponding to the support context vector to the received input context (Paragraphs 0041-0045); and
selecting, as a final output, one of input contexts connected to a plurality of word strings when the next word string is an end word of a sentence (Fig. 2, element 235 “next word”).
Allowable Subject Matter
Claims 6-11 are objected to as being dependent upon a rejected base claim, but would be allowable if rewritten in independent form including all of the limitations of the base claim and any intervening claims.
Conclusion
The prior art made of record and not relied upon is considered pertinent to applicant's disclosure.
Chung et al. (US 2022/0180071) discloses, system and method for adaptive masking and non-directional language understanding and generation.
Chen et al. (US 12,548,559) discloses, Training natural network components.
Contact Information
Any inquiry concerning this communication or earlier communications from the examiner should be directed to Abul K. Azad whose telephone number is (571) 272-7599. If attempts to reach the examiner by telephone are unsuccessful, the examiner's supervisor, Bhavesh Mehta, can be reached at (571) 272-7453.
Any response to this action should be mailed to:
Commissioner for Patents
P.O. Box 1450
Alexandria, VA 22313-1450
Or faxed to: (571) 273-8300.
Hand-delivered responses should be brought to 401 Dulany Street, Alexandria, VA-22314 (Customer Service Window).
Information regarding the status of an application may be obtained from the Patent Application Information Retrieval (PAIR) system. Status information for published applications may be obtained from either Private PAIR or Public PAIR. Status information for unpublished applications is available through Private PAIR only. For more information about the PAIR system, see http://pair-direct.uspto.gov. Should you have questions on access to the Private PAIR system, contact the Electronic Business Center (EBC) at 866-217-9197 (toll-free).
July 27, 2026
/ABUL K AZAD/Primary Examiner, Art Unit 2656
Prosecution Insights