Prosecution Insights
Last updated: November 15, 2026
Application No. 19/661,655

BIOLOGICAL NEURAL NETWORK SYSTEM AND METHODS

Non-Final OA §103§DOUBLEPATENT
Filed
Apr 28, 2026
Priority
Mar 07, 2024 — provisional 63/562,357 +1 more
Examiner
NGUYEN, HENRY K
Art Unit
2121
Tech Center
2100 — Computer Architecture & Software
Assignee
Biological Black Box Inc.
OA Round
1 (Non-Final)
60%
Grant Probability
Moderate
1-2
OA Rounds
88%
With Interview

Examiner Intelligence

Grants 60% of resolved cases
60%
Career Allowance Rate
103 granted / 171 resolved
+5.2% vs TC avg
Strong +28% interview lift
Without
With
+27.7%
Interview Lift
resolved cases with interview
Typical timeline
Avg Prosecution
17 currently pending
Career history
197
Total Applications
across all art units

Office Action

§103 §DOUBLEPATENT
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 . Double Patenting The nonstatutory double patenting rejection is based on a judicially created doctrine grounded in public policy (a policy reflected in the statute) so as to prevent the unjustified or improper timewise extension of the “right to exclude” granted by a patent and to prevent possible harassment by multiple assignees. A nonstatutory double patenting rejection is appropriate where the conflicting claims are not identical, but at least one examined application claim is not patentably distinct from the reference claim(s) because the examined application claim is either anticipated by, or would have been obvious over, the reference claim(s). See, e.g., In re Berg, 140 F.3d 1428, 46 USPQ2d 1226 (Fed. Cir. 1998); In re Goodman, 11 F.3d 1046, 29 USPQ2d 2010 (Fed. Cir. 1993); In re Longi, 759 F.2d 887, 225 USPQ 645 (Fed. Cir. 1985); In re Van Ornum, 686 F.2d 937, 214 USPQ 761 (CCPA 1982); In re Vogel, 422 F.2d 438, 164 USPQ 619 (CCPA 1970); In re Thorington, 418 F.2d 528, 163 USPQ 644 (CCPA 1969). A timely filed terminal disclaimer in compliance with 37 CFR 1.321(c) or 1.321(d) may be used to overcome an actual or provisional rejection based on nonstatutory double patenting provided the reference application or patent either is shown to be commonly owned with the examined application, or claims an invention made as a result of activities undertaken within the scope of a joint research agreement. See MPEP § 717.02 for applications subject to examination under the first inventor to file provisions of the AIA as explained in MPEP § 2159. See MPEP § 2146 et seq. for applications not subject to examination under the first inventor to file provisions of the AIA . A terminal disclaimer must be signed in compliance with 37 CFR 1.321(b). The filing of a terminal disclaimer by itself is not a complete reply to a nonstatutory double patenting (NSDP) rejection. A complete reply requires that the terminal disclaimer be accompanied by a reply requesting reconsideration of the prior Office action. Even where the NSDP rejection is provisional the reply must be complete. See MPEP § 804, subsection I.B.1. For a reply to a non-final Office action, see 37 CFR 1.111(a). For a reply to final Office action, see 37 CFR 1.113(c). A request for reconsideration while not provided for in 37 CFR 1.113(c) may be filed after final for consideration. See MPEP §§ 706.07(e) and 714.13. The USPTO Internet website contains terminal disclaimer forms which may be used. Please visit www.uspto.gov/patent/patents-forms. The actual filing date of the application in which the form is filed determines what form (e.g., PTO/SB/25, PTO/SB/26, PTO/AIA /25, or PTO/AIA /26) should be used. A web-based eTerminal Disclaimer may be filled out completely online using web-screens. An eTerminal Disclaimer that meets all requirements is auto-processed and approved immediately upon submission. For more information about eTerminal Disclaimers, refer to www.uspto.gov/patents/apply/applying-online/eterminal-disclaimer. Claims 21-40 are provisionally rejected on the ground of nonstatutory double patenting as being unpatentable over claim 1-2 and 4-19 of copending Application No. 19054662 (reference application). Although the claims at issue are not identical, they are not patentably distinct from each other because the claims of the ‘662 application are more narrow in scope than the claims of the pending application. This is a provisional nonstatutory double patenting rejection because the patentably indistinct claims have not in fact been patented. Claim 21 is rejected over claim 1 of the ‘662 application. Claim 22 is rejected over claim 2 of the ‘662 application. Claim 23 is rejected over claim 1 of the ‘662 application. Claim 24 is rejected over claim 4 of the ‘662 application. Claim 25 is rejected over claim 5 of the ‘662 application. Claim 26 is rejected over claim 6 of the ‘662 application. Claim 27 is rejected over claim 7 of the ‘662 application. Claim 28 is rejected over claim 8 of the ‘662 application. Claim 29 is rejected over claim 9 of the ‘662 application. Claim 30 is rejected over claim 10 of the ‘662 application. Claim 31 is rejected over claim 11 of the ‘662 application. Claim 32 is rejected over claim 12 of the ‘662 application. Claim 33 is rejected over claim 13 of the ‘662 application. Claim 34 is rejected over claim 14 of the ‘662 application. Claim 35 is rejected over claim 15 of the ‘662 application. Claim 36 is rejected over claim 16 of the ‘662 application. Claim 37 is rejected over claim 17 of the ‘662 application. Claim 38 is rejected over claim 18 of the ‘662 application. Claim 39 is rejected over claim 19 of the ‘662 application. Claim 40 is rejected over claim 1 of the ‘662 application. 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 21-22, 37-40 are rejected under 35 U.S.C. 103 as being unpatentable over Bhasin et al. (US-20250124275-A1) in view of Willett et al. (“A high-performance speech neuroprosthesis”). Regarding Claim 21, Bhasin (US 20250124275 A1) teaches a method for using a dictionary of neural-based embeddings to perform a task using a system comprising: (i) a large language model (LLM) (para [0085] “FIG. 3 illustrates a schematic representation 300 of the architecture of the LLM model such as LLM model 216, in accordance with an embodiment of the present disclosure.”); and (ii) at least one processor, the method comprising: using the system to perform: receiving text comprising one or more tokens to be processed by the LLM in furtherance of performing a task (para [0057] “In an embodiment, the server system 102 is configured to receive a task-specific query from an entity. In various non-limiting examples, the entity may refer to the issuer server 110 associated with an issuing bank, or an acquirer server 108 associated with an acquiring bank. It is noted that the task-specific query may refer to a query from the entity for requesting task-specific information related to one or more suitable products from a plurality of products to be used for performing a task by the entity.” Para [0086] “As may be understood, LLM models are typically built using a transformer architecture. Using the transformer architecture, an LLM model such as LLM model 216 can tokenize an input prompt (such as the task-specific query) into smaller units.”); accessing a dictionary of neural-based embeddings from the memory, the dictionary of neural-based embeddings comprising a respective neural-based embedding for each of the one or more tokens (para [0096] “Pre-processing input prompt: Human language sentences are broken down into words. These words are converted to tokens. In some instances, a dictionary of word to token embeddings is maintained.”); determining, for each token of the one or more tokens, a corresponding neural-based embedding using the dictionary of neural-based embeddings by mapping the respective token to the respective neural-based embedding in the dictionary of neural-based embeddings (para [0093], para [0096], para [0097]), processing the one or more tokens using the LLM by inputting to the LLM the determined corresponding neural-based embeddings for each of the one or more tokens to obtain an output (para [0086] “Further, each token from the plurality of tokens generated for the initial prompt is converted to an embedding. This embedding is shown as input embedding 304 in FIG. 3. As may be understood, these embeddings capture the meaning and context of the various words forming the sentence of the initial prompt.” para [0093] “It is typically used to project the context representations learned by the transformer architecture into a space where each dimension corresponds to a specific token or word in a target vocabulary (or dictionary). The terms “target dictionary” or “target vocabulary” refer to a predefined set of words or tokens that the LLM model 216 is expected to generate as output.”); and using the output in furtherance of performing the task (para [0101] “The LLM model 216 is fine-tuned to perform the task of generating a query response message in response to the task-specific query received from an entity.”). Bhasin does not explicitly disclose the dictionary of neural based embeddings being generated by stimulating a biological neural network (BNN) comprising neurons arranged on a multi-electrode array with stimulation patterns corresponding to respective tokens in a dictionary of tokens; However, Willett (“A high-performance speech neuroprosthesis”) teaches the dictionary of neural based embeddings being generated by stimulating a biological neural network (BNN) comprising neurons arranged on a multi-electrode array with stimulation patterns corresponding to respective tokens in a dictionary of tokens (pg. 1032; “We used two different language models: a large-vocabulary model with 125,000 words (suitable for general English) and a small-vocabulary model with 50 words (suitable for expressing some simple sentences” vocabulary (i.e., dictionary). pg. 1034; “To estimate how each phoneme was neurally represented, we analysed our RNN decoders to extract vectors of neural activity (‘saliency’ vectors) that maximized RNN probability output for each phoneme.” Vector (i.e., word embedding. pg. 1035; “Here we demonstrate a speech BCI that can decode unconstrained sentences from a large vocabulary at a speed of 62 words per minute, using microelectrode arrays to record neural activity at single-neuron resolution.”); Bhasin and Willett are analogous arts directed to the field of neural networks. It would have been obvious to one of ordinary skill in the art before the effective filing date to modify the language model of Bhasin with the multi-electrode array of Willet. Doing so would allow for decoding neural activity evoked by attempted speech into text or sound (abs. Willet). Regarding Claim 22, Bhasin and Willett teach the method of claim 21. Bhasin further teaches further comprising training the LLM using the dictionary of neural-based embeddings (para [0096] “Pre-processing input prompt: Human language sentences are broken down into words. These words are converted to tokens. In some instances, a dictionary of word to token embeddings is maintained.” Para [0097] “Post-processing: Output tokens generated by the LLM model 216 are converted back to the Human language using the same dictionary of word to token embeddings.” Para [0100] “During the pre-training phase, the LLM model 216 is trained on a large generic dataset obtained from various sources. It is understood that the goal of this pre-training phase, the LLM model 216 is allowed to predict the next word in a sentence given the context.”). Regarding Claim 37, Bhasin (US 20250124275 A1) teaches a method for using a dictionary of neural-based embeddings to process text using a large language model (LLM) (para [0055] “In an example, the LLM model 122 is an AI or ML based model that is configured or trained to perform a plurality of operations.”), the LLM having an embedding layer and a plurality of transformers, each of the transformers having one or more attention heads (para [0087] “The addition of positional encoding to the input embedding 304 is depicted by operator 306. Thereafter, the embeddings are fed to a multi-head self-attention mechanism or multi-head attention (see, 308). The multi-head self-attention mechanism allows the LLM model 216 to weigh the importance of different words in the sequence when making predictions.”), the method comprising: using at least one computer hardware processor to perform: generating the dictionary of neural-based embeddings (para [0096] “Pre-processing input prompt: Human language sentences are broken down into words. These words are converted to tokens. In some instances, a dictionary of word to token embeddings is maintained.”) (iv) the least at one processor computer hardware at least in part by: accessing a dictionary of neural-based embeddings from the memory, the dictionary of neural-based embeddings comprising a respective neural-based embedding for each token of the one or more tokens of the text para [0096] “Pre-processing input prompt: Human language sentences are broken down into words. These words are converted to tokens. In some instances, a dictionary of word to token embeddings is maintained.”); determining a corresponding token embedding for each of the one or more tokens by using parameters of the embedding layer to obtain token embeddings, the parameters of the embedding layer determined using the dictionary of neural-based embeddings (para [0088]-[0094] weight (i.e., parameter).), by mapping the respective token to the respective neural-based embedding in the dictionary of neural-based embeddings (para [0093], para [0096], para [0097]), and processing the obtained token embeddings using the plurality of transformers part of the LLM to obtain LLM output (para [0086] “Further, each token from the plurality of tokens generated for the initial prompt is converted to an embedding. This embedding is shown as input embedding 304 in FIG. 3. As may be understood, these embeddings capture the meaning and context of the various words forming the sentence of the initial prompt.” para [0093] “It is typically used to project the context representations learned by the transformer architecture into a space where each dimension corresponds to a specific token or word in a target vocabulary (or dictionary). The terms “target dictionary” or “target vocabulary” refer to a predefined set of words or tokens that the LLM model 216 is expected to generate as output.”). Bhasin does not explicitly disclose generating the dictionary of neural-based embeddings using a biological and artificial neural network (BANN) system comprising (i) a multi-electrode array (MEA); (ii) a biological neural network (BNN) comprising neurons arranged on the MEA; (iii) an artificial neural network (ANN) trained to process multiple features derived from at least one response of the BNN to a stimulation pattern; and the dictionary of neural-based embeddings having been generated by stimulating a biological neural network (BNN) comprising neurons arranged on a multi-electrode array with stimulation patterns corresponding respective tokens in a dictionary of tokens; However, Willett (“A high-performance speech neuroprosthesis”) teaches generating the dictionary of neural-based embeddings using a biological and artificial neural network (BANN) system comprising (i) a multi-electrode array (MEA) (pg. 1031; “To investigate this, we recorded neural activity from four microelectrode arrays— two in area 6v (ventral premotor cortex)10 and two in area 44 (part of Broca’s area)—while our study participant in the BrainGate2 pilot clinical trial attempted to make individual orofacial movements, speak single phonemes or speak single words in response to cues shown on a computer monitor (Fig. 1a,b; Extended Data Fig. 1 shows recorded spike waveforms).”); (ii) a biological neural network (BNN) comprising neurons arranged on the MEA (pg. 1035; “Here we demonstrate a speech BCI that can decode unconstrained sentences from a large vocabulary at a speed of 62 words per minute, using microelectrode arrays to record neural activity at single-neuron resolution.”); (iii) an artificial neural network (ANN) trained to process multiple features derived from at least one response of the BNN to a stimulation pattern (pg. 1035; “Next we investigated how accuracy improved as a function of the number of electrodes used for RNN decoding.”); and the dictionary of neural-based embeddings having been generated by stimulating a biological neural network (BNN) comprising neurons arranged on a multi-electrode array with stimulation patterns corresponding respective tokens in a dictionary of tokens (pg. 1032; “We used two different language models: a large-vocabulary model with 125,000 words (suitable for general English) and a small-vocabulary model with 50 words (suitable for expressing some simple sentences” vocabulary (i.e., dictionary). pg. 1034; “To estimate how each phoneme was neurally represented, we analysed our RNN decoders to extract vectors of neural activity (‘saliency’ vectors) that maximized RNN probability output for each phoneme.” Vector (i.e., word embedding. pg. 1035; “Here we demonstrate a speech BCI that can decode unconstrained sentences from a large vocabulary at a speed of 62 words per minute, using microelectrode arrays to record neural activity at single-neuron resolution.”); Bhasin and Willett are analogous arts directed to the field of neural networks. It would have been obvious to one of ordinary skill in the art before the effective filing date to modify the language model of Bhasin with the multi-electrode array of Willet. Doing so would allow for decoding neural activity evoked by attempted speech into text or sound (abs. Willet). Bhasin and Willett are analogous arts directed to the field of neural networks. It would have been obvious to one of ordinary skill in the art before the effective filing date to modify the language model of Bhasin with the multi-electrode array of Willet. Doing so would allow for decoding neural activity evoked by attempted speech into text or sound (abs. Willet). Regarding Claim 38, Bhasin and Willett teach the method of claim 37. Bhasin further teaches further comprising determining the parameters of the embedding layer using the dictionary of neural-based embeddings, wherein the determining comprises setting the parameters of the embedding layer to the neural-based embeddings (para [0088] “It is noted that the self-attention mechanism computes a weighted sum of all input embeddings where the weights are determined based, at least in part, on the relevance of each word to the current word in the sequence. As may be understood, the self-attention mechanism is configured to compute attention scores using a learned set of parameters (known, as attention weights).”). Regarding Claim 39, Bhasin and Willett teach the method of claim 38. Bhasin further teaches further comprising: training the LLM at least in part by updating the parameters of the embedding layer using further training data, or training the LLM at least in part by fixing the parameters of the embedding layer during training and updating other parameters of the LLM using further training data (para [0090] “It is noted that layer normalization and residual connections are used by the transformer architecture to stabilize the training of the LLM model 216 thereby, enabling the LLM model 216 to learn effectively. It is understood that residual connections allow gradients to flow more easily through the neural network.”). Regarding claim 40, Claim 40 is the system corresponding to the method of claim 21. Claim 40 is substantially similar to claim 21 and is rejected on the same grounds. Conclusion Any inquiry concerning this communication or earlier communications from the examiner should be directed to HENRY K NGUYEN whose telephone number is (571)272-0217. The examiner can normally be reached Mon - Fri 9:00am-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, Li B Zhen can be reached at 5712723768. 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. /HENRY NGUYEN/Primary Examiner, Art Unit 2121
Read full office action

Prosecution Timeline

Apr 28, 2026
Application Filed
Oct 20, 2026
Non-Final Rejection mailed — §103, §DOUBLEPATENT (current)

Precedent Cases

Applications granted by this same examiner with similar technology

Patent 12793402
MACHINE LEARNING METHODS AND APPARATUS RELATED TO PREDICTING MOTION(S) OF OBJECT(S) IN A ROBOT'S ENVIRONMENT BASED ON IMAGE(S) CAPTURING THE OBJECT(S) AND BASED ON PARAMETER(S) FOR FUTURE ROBOT MOVEMENT IN THE ENVIRONMENT
5y 0m to grant Granted Nov 10, 2026
Patent 12783434
TRANSFER DATA IN A MEMORY SYSTEM WITH ARTIFICIAL INTELLIGENCE MODE
7y 2m to grant Granted Nov 03, 2026
Patent 12748946
Method, System, and Computer Program Product for Efficient Content-Based Time Series Retrieval
1y 9m to grant Granted Sep 29, 2026
Patent 12718058
HYBRID NEURAL NETWORK ARCHITECTURE WITHIN CASCADING PIPELINES
5y 8m to grant Granted Aug 25, 2026
Patent 12705302
METHOD, ACCELERATOR, AND ELECTRONIC DEVICE WITH TENSOR PROCESSING
5y 9m to grant Granted Aug 11, 2026
Study what changed to get past this examiner. Based on 5 most recent grants.

Strategy Recommendation AI-generated — please review before filing

Get a prosecution strategy drawn from examiner precedents, rejection analysis, and claim mapping.
Typically takes 5-10 seconds — AI-generated, attorney review required before filing

Prosecution Projections

1-2
Expected OA Rounds
60%
Grant Probability
88%
With Interview (+27.7%)
Low
PTA Risk
Based on 171 resolved cases by this examiner. Grant probability derived from career allowance rate.

Sign in with your work email

Enter your email to receive a magic link. No password needed.

Personal email addresses (Gmail, Yahoo, etc.) are not accepted.

Free tier: 3 strategy analyses per month