Enterprise generative artificial intelligence anti-hallucination and … — C3.Ai, Inc. (US20240370709A1)

ABSTRACT

Abstract

An anti-hallucination and attribution architecture for enterprise generative AI systems is disclosed herein which increases the accuracy and reliability of generative artificial intelligence content (e.g., responses or answers) by detecting, preventing, and mitigating hallucination. The anti-hallucination and attribution architecture can be added to deployed generative artificial intelligence systems as a separate tool or module, which allows it to work with the deployed systems without having to retool or redesign those systems. The anti-hallucination and attribution architecture can also be deployed with minimal impact on live production systems.

Description

CROSS-REFERENCE TO RELATED APPLICATIONS

The present application claims the benefit of U.S. Provisional Patent Application Ser. No. 63/499,485 filed May 1, 2023, and entitled “Systems and Methods for Parallelized Model Management,” which is hereby incorporated by reference herein. The present application is a continuation-in-part of U.S. Nonprovisional patent application Ser. No. 18/542,481 filed Dec. 15, 2023, and entitled “Generative Artificial Intelligence Enterprise Search,” U.S. Nonprovisional patent application Ser. No. 18/542,536 filed Dec. 15, 2023, and entitled “Enterprise Generative Artificial intelligence Architecture,” and U.S. Nonprovisional patent application Ser. No. 18/542,583 filed Dec. 15, 2023, and entitled “Generative Artificial Intelligence Crawling and Chunking,” each of which is hereby incorporated by reference herein.

TECHNICAL FIELD

This disclosure pertains to generative artificial intelligence and machine learning. More specifically, this disclosure pertains to anti-hallucination and attribution architectures for enterprise generative artificial intelligence.

BACKGROUND

Generative artificial intelligence (generative AI) refers to a subfield of machine learning concerned with algorithms that can generate new data instances. These algorithms are typically deep learning models trained on large datasets to learn the underlying statistical properties of the data. Generative artificial intelligence approaches employ artificial neural networks to mimic statistical properties of large training data. Different from conventional AI approaches, which might focus on analyzing existing data or completing specific tasks based on pre-defined rules, Generative AI can create entirely new output using its learned understanding of the data. Conventional generative AI approaches suffer from a variety of drawbacks, such as hallucination and an inability for users to verify or corroborate generative artificial intelligence responses.

BRIEF DESCRIPTION OF THE DRAWINGS

FIG. 1 depicts a diagram of an example logical flow of an enterprise generative artificial intelligence system with an anti-hallucination and attribution architecture according to some embodiments.

FIG. 2 depicts a diagram of an example parsing and information graph generation process for enterprise generative artificial intelligence system using an anti-hallucination and attribution architecture according to some embodiments.

FIG. 3 depicts a flowchart of an example anti-hallucination and attribution process for enterprise generative artificial intelligence systems according to some embodiments.

FIG. 4 depicts a diagram of an example structure of response segments generated by an anti-hallucination and attribution method for enterprise generative artificial intelligence systems according to some embodiments.

FIG. 5 depicts a diagram of an example enterprise generative artificial intelligence system architecture and environment according to some embodiments.

FIG. 6 A depicts an example enterprise search graphical user interface and underlying architecture according to some embodiments.

FIG. 6 B depicts an example enterprise generative artificial intelligence response graphical user interface according to some embodiments.

FIG. 7 depicts a diagram of an example layered architecture and environment of an enterprise generative artificial intelligence system according to some embodiments.

FIG. 8 depicts a diagram of an example network system for enterprise generative artificial intelligence according to some embodiments.

FIG. 9 depicts a flowchart of an example iterative generative artificial intelligence process using unstructured data according to some embodiments.

FIG. 10 A depicts a flowchart of an example iterative generative artificial intelligence process using unstructured data according to some embodiments.

FIG. 10 B depicts a flowchart of an example non-iterative generative artificial intelligence process using unstructured data according to some embodiments.

FIG. 10 C depicts a flowchart of an example non-iterative generative artificial intelligence process using unstructured data according to some embodiments.

FIG. 11 depicts a flowchart of an example iterative generative artificial intelligence process using unstructured data according to some embodiments.

FIG. 12 depicts a flowchart of an example enterprise generative artificial intelligence method according to some embodiments.

FIG. 13 depicts a flowchart of an example enterprise generative artificial intelligence method using an agent-orchestrator architecture according to some embodiments.

FIG. 14 depicts a flowchart of an example anti-hallucination and attribution method for enterprise generative artificial intelligence systems according to some embodiments.

FIG. 15 depicts a flowchart of an example information retrieval method for enterprise generative artificial intelligence method according to some embodiments.

FIG. 16 depicts a flowchart of an example method of routing requests to different agent programs and validating responses for enterprise generative artificial according to some embodiments.

FIG. 17 depicts a diagram of an example computer system for implementing the features disclosed herein according to some embodiments.

DETAILED DESCRIPTION

Generative AI is an artificial intelligence technology that uses machine learning algorithms to perform tasks that imitate human cognitive intelligence and generate content. Content can be in the form of text, audio, video, images, and more. However, conventional generative artificial intelligence processes often present biased or faulty information as a result of machine learning model hallucination (or, simply, hallucination). In generative AI, an artificial hallucination, confabulation, or delusion, refers to a specific type of output generated by the model that deviates from factual accuracy, ground truth, or the intended context. Conventional generative AI can create content that does not correspond to reality or verifiable information; commonly contributed to by statistical inaccuracy, model overfitting, and data bias. The generated content may appear plausible and be internally coherent but lack accuracy and coherence with the provided prompt or surrounding information. Enterprise environments can also include contradictory information, which further increases the likelihood of hallucination. Hallucination in enterprise environments can is complicated by often incomplete and disparate information that is spread across a variety of different, and incompatible, enterprise systems. Currently generative Al systems fail to detect, prevent, or mitigate hallucination. Conventional generative artificial intelligence processes also fail to provide any mechanism for users to verify, validate, or corroborate generative artificial intelligence content. Accordingly, users have no way of knowing if generative artificial intelligence content is accurate or the result of hallucination.

An anti-hallucination and attribution architecture for enterprise generative AI systems is disclosed herein which increases the accuracy and reliability of generative artificial intelligence content (e.g., responses or answers) by detecting, preventing, and mitigating hallucination. Moreover, the anti-hallucination and attribution architecture can be added to deployed generative artificial intelligence systems as a separate tool or module, which allows it to work with the deployed systems without having to retool or redesign those systems. The anti-hallucination and attribution architecture can also be deployed with minimal impact on live production systems.

The enterprise generative artificial intelligence systems described herein can transform interactions with enterprise information that fundamentally changes the human-computer interaction (HCI) model for enterprise software. Enterprises running sensitive workloads in cloud-native, on premise, or air-gapped environments can implement enterprise generative Al architectures to generate enterprise-wide insights using tools to rapidly locate and retrieve information with agents that develop and coordinate complex operations in response to simple intuitive input. The enterprise generative AI architecture enables enterprise users to ask open-ended, multi-level, context specific questions that are processed used generative AI with machine learning to understand the request, identify relevant information, and generate new context specific insights with predictive analysis. The enterprise generative AI architecture supports simplified human-computer-interactions with intuitive natural language interface as well as advanced accessibility features for adaptable forms of input including but not limited to text, audio, video, images, and more.

The anti-hallucination and attribution architecture includes an anti-hallucination and attribution module that can work with deployed enterprise generative artificial intelligence systems and components (e.g., retriever tools, large language models or other generative artificial intelligence models, etc.) in order to detect, prevent, and mitigate hallucination caused by the deployed large language models or other models (e.g., generative artificial intelligence models, multimodal models). For example, a user can submit a prompt (e.g., question, query, etc.) to the enterprise generative artificial intelligence system, such as “How many different engineers has John Doe worked with within his engineering department?” This may require the enterprise generative artificial intelligence system to identify John Doe, identify John Doe's department, determine the engineers in that department in a third iteration, determine which of those engineers John Doe has interacted with, and then finally combine those results, to generate the answer to the query. Such complex queries can introduce hallucination at any of step along the way to generating the answer. For example, there may be several John Doe's within an organization, and the system could easily hallucinate in order to determine which John Doe the query is referring to. The anti-hallucination and attribution architecture can prevent or mitigate such hallucination by providing the answer to an anti-hallucination and attribution module prior to generating a final answer.

The anti-hallucination and attribution module can parse the answer generated by the enterprise generative artificial intelligence system into several chunks (e.g., several sentences). The anti-hallucination and attribution module can process the chunks along with the original passages retrieved by the enterprise generative artificial intelligence system in order to find relevant passages for each chunk. The anti-hallucination and attribution module can then combine the chunks and the retrieved relevant passages to generate an attributed answer. For example, the attributed answer may include source identifiers that indicate the documents and passages that were used to generate the answer. Accordingly, the system provides traceable attribution to confirm that the answer is reliable, and the model was not hallucinating. If the anti-hallucination and attribution module cannot locate any relevant passages, the anti-hallucination and attribution module may determine that the model hallucinated and either re-run the query to find another result or notify the user that a reliable result could not be determined, rather than merely providing the response to the user.

The enterprise generative artificial intelligence systems described herein can further use a combination of agents and tools to efficiently process a wide variety of inputs received from disparate data sources (e.g., having different data formats) and return results in a common data format (e.g., natural language). The enterprise generative artificial intelligence architecture includes an orchestrator agent (or, simply, orchestrator) that supervises, controls, and/or otherwise administrates many different agents, tools, and/or modules (e.g., an anti-hallucination and attribution module). Orchestrators can include one or more machine learning models and can execute supervisory functions, such as routing inputs (e.g., queries, instruction sets, natural language inputs or other human-readable inputs, machine-readable inputs) to specific agents to accomplish a set of prescribed tasks (e.g., retrieval requests prescribed by the orchestrator to answer a query). Machine learning models can include some or all of the different types or modalities of models described herein (e.g., multimodal machine learning models, large language models, data models, statistical models, audio models, visual models, audiovisual models, etc.). Agents can include one or more multimodal models (e.g., large language models) to accomplish the prescribed tasks using a variety of different tools. Different agents can use various tools to execute and process unstructured data retrieval requests, structured data retrieval requests, API calls (e.g., for accessing artificial intelligence application insights), and the like. Tools can include one or more specific functions and/or machine learning models to accomplish a given task (or set of tasks).

Agents can adapt to perform differently based on contexts. A context may relate to a particular domain (e.g., industry) and an agent may employ a particular model (e.g., large language model, other machine learning model, and/or data model) that has been trained on industry-specific datasets, such as healthcare datasets. The particular agent can use a healthcare model when receiving inputs associated with a healthcare environment and can also easily and efficiently adapt to use a different model based on different inputs or context. Indeed, some or all of the models described herein may be trained for specific domains in addition to, or instead of, more general purposes. The enterprise generative artificial intelligence architecture leverages domain specific models to produce accurate context specific retrieval and insights.

The orchestrator manages the agents to efficiently process disparate inputs or different portions of an input. For example, an input may require the system to access and retrieve data records from d

CROSS-REFERENCE TO RELATED APPLICATIONS

The present application claims the benefit of U.S. Provisional Patent Application Ser. No. 63/499,485 filed May 1, 2023, and entitled “Systems and Methods for Parallelized Model Management,” which is hereby incorporated by reference herein. The present application is a continuation-in-part of U.S. Nonprovisional patent application Ser. No. 18/542,481 filed Dec. 15, 2023, and entitled “Generative Artificial Intelligence Enterprise Search,” U.S. Nonprovisional patent application Ser. No. 18/542,536 filed Dec. 15, 2023, and entitled “Enterprise Generative Artificial intelligence Architecture,” and U.S. Nonprovisional patent application Ser. No. 18/542,583 filed Dec. 15, 2023, and entitled “Generative Artificial Intelligence Crawling and Chunking,” each of which is hereby incorporated by reference herein.

TECHNICAL FIELD

This disclosure pertains to generative artificial intelligence and machine learning. More specifically, this disclosure pertains to anti-hallucination and attribution architectures for enterprise generative artificial intelligence.

BACKGROUND

Generative artificial intelligence (generative AI) refers to a subfield of machine learning concerned with algorithms that can generate new data instances. These algorithms are typically deep learning models trained on large datasets to learn the underlying statistical properties of the data. Generative artificial intelligence approaches employ artificial neural networks to mimic statistical properties of large training data. Different from conventional AI approaches, which might focus on analyzing existing data or completing specific tasks based on pre-defined rules, Generative AI can create entirely new output using its learned understanding of the data. Conventional generative AI approaches suffer from a variety of drawbacks, such as hallucination and an inability for users to verify or corroborate generative artificial intelligence responses.

BRIEF DESCRIPTION OF THE DRAWINGS

FIG. 1 depicts a diagram of an example logical flow of an enterprise generative artificial intelligence system with an anti-hallucination and attribution architecture according to some embodiments.

FIG. 2 depicts a diagram of an example parsing and information graph generation process for enterprise generative artificial intelligence system using an anti-hallucination and attribution architecture according to some embodiments.

FIG. 3 depicts a flowchart of an example anti-hallucination and attribution process for enterprise generative artificial intelligence systems according to some embodiments.

FIG. 4 depicts a diagram of an example structure of response segments generated by an anti-hallucination and attribution method for enterprise generative artificial intelligence systems according to some embodiments.

FIG. 5 depicts a diagram of an example enterprise generative artificial intelligence system architecture and environment according to some embodiments.

FIG. 6 A depicts an example enterprise search graphical user interface and underlying architecture according to some embodiments.

FIG. 6 B depicts an example enterprise generative artificial intelligence response graphical user interface according to some embodiments.

FIG. 7 depicts a diagram of an example layered architecture and environment of an enterprise generative artificial intelligence system according to some embodiments.

FIG. 8 depicts a diagram of an example network system for enterprise generative artificial intelligence according to some embodiments.

FIG. 9 depicts a flowchart of an example iterative generative artificial intelligence process using unstructured data according to some embodiments.

FIG. 10 A depicts a flowchart of an example iterative generative artificial intelligence process using unstructured data according to some embodiments.

FIG. 10 B depicts a flowchart of an example non-iterative generative artificial intelligence process using unstructured data according to some embodiments.

FIG. 10 C depicts a flowchart of an example non-iterative generative artificial intelligence process using unstructured data according to some embodiments.

FIG. 11 depicts a flowchart of an example iterative generative artificial intelligence process using unstructured data according to some embodiments.

FIG. 12 depicts a flowchart of an example enterprise generative artificial intelligence method according to some embodiments.

FIG. 13 depicts a flowchart of an example enterprise generative artificial intelligence method using an agent-orchestrator architecture according to some embodiments.

FIG. 14 depicts a flowchart of an example anti-hallucination and attribution method for enterprise generative artificial intelligence systems according to some embodiments.

FIG. 15 depicts a flowchart of an example information retrieval method for enterprise generative artificial intelligence method according to some embodiments.

FIG. 16 depicts a flowchart of an example method of routing requests to different agent programs and validating responses for enterprise generative artificial according to some embodiments.

FIG. 17 depicts a diagram of an example computer system for implementing the features disclosed herein according to some embodiments.

DETAILED DESCRIPTION

Generative AI is an artificial intelligence technology that uses machine learning algorithms to perform tasks that imitate human cognitive intelligence and generate content. Content can be in the form of text, audio, video, images, and more. However, conventional generative artificial intelligence processes often present biased or faulty information as a result of machine learning model hallucination (or, simply, hallucination). In generative AI, an artificial hallucination, confabulation, or delusion, refers to a specific type of output generated by the model that deviates from factual accuracy, ground truth, or the intended context. Conventional generative AI can create content that does not correspond to reality or verifiable information; commonly contributed to by statistical inaccuracy, model overfitting, and data bias. The generated content may appear plausible and be internally coherent but lack accuracy and coherence with the provided prompt or surrounding information. Enterprise environments can also include contradictory information, which further increases the likelihood of hallucination. Hallucination in enterprise environments can is complicated by often incomplete and disparate information that is spread across a variety of different, and incompatible, enterprise systems. Currently generative Al systems fail to detect, prevent, or mitigate hallucination. Conventional generative artificial intelligence processes also fail to provide any mechanism for users to verify, validate, or corroborate generative artificial intelligence content. Accordingly, users have no way of knowing if generative artificial intelligence content is accurate or the result of hallucination.

An anti-hallucination and attribution architecture for enterprise generative AI systems is disclosed herein which increases the accuracy and reliability of generative artificial intelligence content (e.g., responses or answers) by detecting, preventing, and mitigating hallucination. Moreover, the anti-hallucination and attribution architecture can be added to deployed generative artificial intelligence systems as a separate tool or module, which allows it to work with the deployed systems without having to retool or redesign those systems. The anti-hallucination and attribution architecture can also be deployed with minimal impact on live production systems.

The enterprise generative artificial intelligence systems described herein can transform interactions with enterprise information that fundamentally changes the human-computer interaction (HCI) model for enterprise software. Enterprises running sensitive workloads in cloud-native, on premise, or air-gapped environments can implement enterprise generative Al architectures to generate enterprise-wide insights using tools to rapidly locate and retrieve information with agents that develop and coordinate complex operations in response to simple intuitive input. The enterprise generative AI architecture enables enterprise users to ask open-ended, multi-level, context specific questions that are processed used generative AI with machine learning to understand the request, identify relevant information, and generate new context specific insights with predictive analysis. The enterprise generative AI architecture supports simplified human-computer-interactions with intuitive natural language interface as well as advanced accessibility features for adaptable forms of input including but not limited to text, audio, video, images, and more.

The anti-hallucination and attribution architecture includes an anti-hallucination and attribution module that can work with deployed enterprise generative artificial intelligence systems and components (e.g., retriever tools, large language models or other generative artificial intelligence models, etc.) in order to detect, prevent, and mitigate hallucination caused by the deployed large language models or other models (e.g., generative artificial intelligence models, multimodal models). For example, a user can submit a prompt (e.g., question, query, etc.) to the enterprise generative artificial intelligence system, such as “How many different engineers has John Doe worked with within his engineering department?” This may require the enterprise generative artificial intelligence system to identify John Doe, identify John Doe's department, determine the engineers in that department in a third iteration, determine which of those engineers John Doe has interacted with, and then finally combine those results, to generate the answer to the query. Such complex queries can introduce hallucination at any of step along the way to generating the answer. For example, there may be several John Doe's within an organization, and the system could easily hallucinate in order to determine which John Doe the query is referring to. The anti-hallucination and attribution architecture can prevent or mitigate such hallucination by providing the answer to an anti-hallucination and attribution module prior to generating a final answer.

The anti-hallucination and attribution module can parse the answer generated by the enterprise generative artificial intelligence system into several chunks (e.g., several sentences). The anti-hallucination and attribution module can process the chunks along with the original passages retrieved by the enterprise generative artificial intelligence system in order to find relevant passages for each chunk. The anti-hallucination and attribution module can then combine the chunks and the retrieved relevant passages to generate an attributed answer. For example, the attributed answer may include source identifiers that indicate the documents and passages that were used to generate the answer. Accordingly, the system provides traceable attribution to confirm that the answer is reliable, and the model was not hallucinating. If the anti-hallucination and attribution module cannot locate any relevant passages, the anti-hallucination and attribution module may determine that the model hallucinated and either re-run the query to find another result or notify the user that a reliable result could not be determined, rather than merely providing the response to the user.

The enterprise generative artificial intelligence systems described herein can further use a combination of agents and tools to efficiently process a wide variety of inputs received from disparate data sources (e.g., having different data formats) and return results in a common data format (e.g., natural language). The enterprise generative artificial intelligence architecture includes an orchestrator agent (or, simply, orchestrator) that supervises, controls, and/or otherwise administrates many different agents, tools, and/or modules (e.g., an anti-hallucination and attribution module). Orchestrators can include one or more machine learning models and can execute supervisory functions, such as routing inputs (e.g., queries, instruction sets, natural language inputs or other human-readable inputs, machine-readable inputs) to specific agents to accomplish a set of prescribed tasks (e.g., retrieval requests prescribed by the orchestrator to answer a query). Machine learning models can include some or all of the different types or modalities of models described herein (e.g., multimodal machine learning models, large language models, data models, statistical models, audio models, visual models, audiovisual models, etc.). Agents can include one or more multimodal models (e.g., large language models) to accomplish the prescribed tasks using a variety of different tools. Different agents can use various tools to execute and process unstructured data retrieval requests, structured data retrieval requests, API calls (e.g., for accessing artificial intelligence application insights), and the like. Tools can include one or more specific functions and/or machine learning models to accomplish a given task (or set of tasks).

Agents can adapt to perform differently based on contexts. A context may relate to a particular domain (e.g., industry) and an agent may employ a particular model (e.g., large language model, other machine learning model, and/or data model) that has been trained on industry-specific datasets, such as healthcare datasets. The particular agent can use a healthcare model when receiving inputs associated with a healthcare environment and can also easily and efficiently adapt to use a different model based on different inputs or context. Indeed, some or all of the models described herein may be trained for specific domains in addition to, or instead of, more general purposes. The enterprise generative artificial intelligence architecture leverages domain specific models to produce accurate context specific retrieval and insights.

The orchestrator manages the agents to efficiently process disparate inputs or different portions of an input. For example, an input may require the system to access and retrieve data records from disparate data sources (e.g., unstructured datastores, structured datastores, timeseries datastores, and the like), database tables from different types of databases, and machine learning insights from different machine learning applications. The different agents can each separately, and in parallel, handle each of these requests, greatly increasing computational efficiency.

Agents can process the disparate data returned by the different agents and/or tools. For example, large language models typically receive inputs in natural language format. The agents may receive information in a non-natural language format (e.g., database table, image, audio) from a tool and transform it into natural language describing the tool output in a format understood by large language models. A model (e.g., large language model, multimodal model) can then process that input to generate an initial answer, which the anti-hallucination and attribution module can validate and/or attribute prior to providing a final output.

FIG. 1 depicts a diagram 100 of an example logical flow of an enterprise generative artificial intelligence system with an anti-hallucination and attribution architecture according to some embodiments. Generally, generative artificial intelligence models (e.g., large language models, multimodal models) have different capabilities and follow instructions with varying levels of performance. In one example, an enterprise generative artificial intelligence system can generate attributions (e.g., source citations) for responses generated by the enterprise generative artificial intelligence system (e.g., by a large language model) through prompting the generative artificial intelligence model to cite its sources. However, this approach has several drawbacks and limitations.

More specifically, different models may be able to follow this instruction or not, and this can depend on the particular model used by the enterprise generative artificial intelligence system, as well as the task or query that is received. The formatting of the source hash definition used by the enterprise generative artificial intelligence system to obtain citations can be different from customer to customer which in turn affects the citation performance. This approach can also limit the system's ability to cite only from the initially retrieved documents. This can potentially cause the system to ignore important or relevant documents (e.g., because the system did not pick a sufficiently high enough value for the number of retrieved documents which may vary from query to query) and prevents the system from reaping the potential benefits of leveraging a fine-tuned model on a customer corpus associated with enterprise generative artificial intelligence system.

Moreover, while this attribution approach may reduce hallucination to some degree, it still does not fully or optimally prevent hallucinations. Since generating attributions and preventing hallucinations are extremely important for enterprise generative artificial intelligence systems, it cannot be left to chance that the model would follow the instructions (e.g., instructions to cite sources). Indeed, the enterprise generative artificial intelligence systems described herein may be model-agnostic, and some models may implement such an approach very poorly. That is, this attribution approach is compatible with different types of models include proprietary or open source large language models (LLMs), small language models, image generation models, audio generation models, video generation models, omni-modal models, etc. Accordingly, an approach is needed that not only prevents hallucinations but is also model-agnostic.

For example, rather than prompting the model to cite its sources, the enterprise generative artificial intelligence system performs attribution as a separate step after the model has generated its response. This can address the limitations discussed above and also allows the system while using information in retrieved passages to also rely on the information encoded in a fine-tuned model on their corpus. The system can combine local and global search in the corpus to provide better and more informative responses. This enables the system to use this new component (e.g., anti-hallucination and attribution module 110 ) as a separate tool which can be used to fact check statements and provide corroborating statements.

An overview of the current architecture for anti-hallucination and attribution is shown in FIG. 1 . In the example of FIG. 1 , a query 102 is received (e.g., a question received from a user through a graphical user interface). The query 102 is passed to retriever 104 to retrieve passages (e.g., from a vector store) that are relevant to the query (e.g., using a similarity evaluation that generates a corresponding similarity score). The retriever 104 passes the retrieved passages through a context processor 106 that forms a prompt that is then provided to the generative artificial intelligence model 108 for generating the response. In the new design, rather than merely providing the response to the user, the system applies the anti-hallucination and attribution module 110 to this response. The response can initially go through a response parser 112 that parses the response into chunks (or, segments) to be attributed. These chunks and the original retrieved passages are processed by the anti-hallucination and attribution (AHA) retriever 114 to locate and retrieve (e.g., from a vector store) relevant passages for each chunk. The chunks together with their corresponding retrieved passages arc then provided to the response renderer 116 that combines the corresponding retrieved passages and the chunks to generate the attributed response 118 .

The anti-hallucination and attribution module 110 can parse and chunk the response. To this end, the anti-hallucination and attribution module 110 can break up, segment, or select a subset of the response into sentences and contiguous pieces. These pieces or portions are then combined by the anti-hallucination and attribution module 110 into bigger chunks until it reaches a maximum number of tokens in each chunk. For computing the tokens, the anti-hallucination and attribution module 110 can utilize a tokenizer chosen to be the same tokenizer used by the embedding model in the anti-hallucination and attribution retriever 114 . Each of the chunks are then fed through anti-hallucination and attribution retriever 114 to generate relevant passages to each chunk. Furthermore, in this stage the anti-hallucination and attribution module 110 has the option to limit or filter the passages used for attribution to the original passages retrieved in the beginning of the pipeline (e.g., by setting a tag value). These passages together with the chunks are then passed to the response renderer 116 to generate the attributed response 118 . In this stage, the anti-hallucination and attribution module 110 has the option to include several passages for supporting each chunk. This can be controlled through a parameter value. For example, if set to a value larger than 1 the anti-hallucination and attribution module 110 can cite more passages for each chunk only if their score is larger than a certain threshold value of the anti-hallucination and attribution retriever 114 . The anti-hallucination and attribution module 110 can always include at least one citation for each chunk, though the anti-hallucination and attribution module 110 may only include more if the score associated to the passage is larger than the threshold. The anti-hallucination and attribution module 110 can also provide a visualization such as a color code based on the maximum score of the cited passages for the chunk.

In one example, an original response generated by the generative artificial intelligence model 108 can include the following:

Parcel B1 is an approximately 40-50 acre property and adjacent wharf infrastructure purpose-built for marshalling (with dredged berth pockets and turning basin) (“Parcel B1”). The development of Parcel A is targeted to be completed in Quarter 2 of 2024, with the completion of the development of the balance of Phase 1 anticipated in 2024 and 2025. Development of Phase 2 is targeted to commence in 2024 and to achieve completion by 2028. The costs of development of Phases 1 and 2 are estimated at approximately $550 million each, excluding hydraulic dredging undertaken by the Department of Transportation. The Authority is issuing bonds to cover some of these costs.

The corresponding attributed response generated by the anti-hallucination and attribution module 110 can include the following:

Parcel A is targeted to be completed in Quarter 2 of 2024, with the completion of the development of the balance of Phase 1 anticipated in 2024 and 2025. Development of Phase 2 is targeted to commence in 2024 and to achieve completion by 2028 (from with score 0.7580196261405945). The costs of development of Phases 1 and 2 are estimated at approximately $550 million each, excluding hydraulic dredging undertaken by the Department of Transportation. The Authority is issuing bonds to cover some of these costs (from with score 0.8340072631835938). Parcel B1 is an approximately 40-50 acre property and adjacent wharf infrastructure purpose-built for marshalling (with dredged berth pockets and turning basin) (“Parcel B1”) (from with score 0.7503519654273987). The development of Parcel A is targeted to be completed in Quarter 2 of 2024, with the completion of the development of the balance of Phase 1 anticipated in 2024 and 2025. Development of Phase 2 is targeted to commence in 2024 and to achieve completion by 2028 (from with score 0.7610742449760437).

In some embodiments, the anti-hallucination and attribution module 110 can visually mark-up (e.g., highlight, color-code) the attributed response. For example, the color yellow may be used highlight chunks (e.g., sentences) that have a relatively low similarity score (e.g., relative to a threshold value). The color red may be used to highlight chunks for which no relevant passages were located by the anti-hallucination and attribution retriever 114 . The color green may be used to highlight chunks that have a relatively high similarity score (e.g., relative to the threshold value) indicating that chunk was not generated by any hallucinations of the generative artificial intelligence model 108 .

In another example, an original response generated by the generative artificial intelligence model 108 can include:

No, you do not need to pay federal tax and state tax for interest earned on the 2023 Series A Bonds, as they are not excluded from gross income for federal income tax purposes and are not included in gross income under the existing Gross Income Tax Act. However, it is recommended to consult with a tax professional for further guidance.

The corresponding attributed response generated by the anti-hallucination and attribution module 110 can include the following:

No, you do not need to pay federal tax and state tax for interest earned on the 2023 Series A Bonds, as they are not excluded from gross income for federal income tax purposes and are not included in gross income under the existing Gross Income Tax Act (from with score 0.7880104780197144). However, it is recommended to consult with a tax professional for further guidance (from with score 0.4502568244934082).

FIG. 2 depicts a diagram 200 of an example parsing and information graph generation process for enterprise generative artificial intelligence system using an anti-hallucination and attribution architecture according to some embodiments. The example data record preprocessing and information retrieval process may be performed by one or more of the systems and/or subsystems described herein (e.g., enterprise generative artificial intelligence system 802 ). The information graph may be used to facilitate information retrieval in order to accurately attribute source passages to chunks of a response and prevent hallucination.

Generally, data records can include different modalities of information, such as pure text, tables, images, code, video, audio, and the like. In order to effectively and reliably retrieve information for a query, the preprocessing and information retrieval process provides a multimodal approach for extracting information from the data records.

The preprocessing and information retrieval process may include three stages. The first stage may include parsing and extracting different modalities from these documents. This process can be done in parallel (or substantially in parallel) for all different modalities.

In one example, text and code can be parsed (step 202 ). Depending on the file format, extracting textual information, that is pure text and code, can come with different difficulties and may require leveraging different libraries. Irrespective of the data record type, the process can go through various steps to prepare for other downstream stages. Depending on the file format, some of these steps may increase in complexity.

One of these steps can include extracting textual information (step 204 ) from data records 201 so that they can be further parsed. For example, extracting everything that is not an image or table. The output of this step can include (e.g., may be required to include) all the textual information (e.g., information that are not image and table captions) from the data records, which can then be used for further separating text and code pieces. The parsed results may have a high fidelity (e.g., no introduction of random spaces, weird characters that would break the meaning of sentences), and robust to font, size, colors, location with respect to other elements of the pages).

Another step can include separating text and code (step 205 ). The goal of this step is to identify and separate code and pure text. This then allows the system to further process these modalities separately. An additional step can include chunking and parsing text and code (step 206 ). Having separated the text and code modalities, the purpose of this step is to identify, locate and extract contiguous pieces of code (step 207 ), and conduct chunking of the text contents (step 206 ) in reasonable and as contiguous fashion (e.g., that is no cutting mid-sentence or mid-paragraph especially due to breakages between pages and if possible, having chunks with contiguous topics).

The system (e.g., enterprise generative artificial intelligence system 802 ) may use an object-oriented structure, where there may be classes for pure text and code. These classes, at a minimum, can have fields keeping track of the content, location in the document, number of tokens of the content (e.g., that means that the system should have knowledge of the tokenizer for this purpose). For the text class, the system may already be able to keep track of what code snippets have been removed from the chunk's content (or are related to it). This may have already been done as part of the system.

In another example, tables can be parsed. An important modality in data records is tables. To be able to conduct effective information retrieval, the system can first locate and identify tables in full (e.g., since tables may span multiple pages/data records/segments or may appear in different structures in a single page). See step 206 . The system can identify libraries that enable these features and measure their performance in identifying tables in full. Once the tables are identified, the tables may be extracted as an image or as a dataframe, and the like (step 209 ). The system can also extract the caption or title of the table, column headers, potentially also the row index(es), etc., as metadata associated with it (step 209 ). Similar to text and code classes, the system can also have a table class that keeps track of the extracted table content, its location, title/caption, etc.

In another example, images can be parsed. Similar to how the system handles tables, the system can also start with identifying and locating images in full (step 212 ). The system can include an image class to keep track of the images in the documents, that is the content, the location of the image in the doc and its caption/title as metadata associated with it. The system can also extract images as metadata associated with it (step 213 ).

At the end of the first stage, the system can have several instances of the text, code, table and image classes, outlining different modalities in each data record. Having done this, the system can continue to the second stage, namely, building an information graph for each data record.

In the second stage, in order to facilitate an effective information retrieval, the system can represent the information in each data record 201 by generating and/or using an information graph 220 . The

nodes

230 , 241 , 243 and 245 of this graph correspond to different instances of the four classes of modality from stage one. The depiction is a (directed) bipartite graph, with edges going out from the text nodes to all other modality nodes. Establishing these edges are the primary goal of this stage.

In the example of FIG. 2 , there is an edge between a text node 230 to other modality nodes

241 , 243 , and/or 245 , if there is reference (e.g., relation) between them. This can be established based on direct references to them in the text chunk, based on proximity or even based on contextual similarity with their caption or title. Having identified the edges, the system can keep track of what edges are between each text node 230 and other modality nodes

241 , 243 , and/or 245 as part of the text class. Once the system has fully specified this graph, the system can design an information retrieval process leveraging this graph 220 . For example, agents (e.g., agents 906 ) and tools (e.g., tools 908 ) can use the graph to retrieve information. In another example, the anti-hallucination and attribution module 110 can use the graph 220 to attribute sources to a response to generate an attributed response. To this end, the information graph 220 may also include access control protocol information (e.g., as defined by the enterprise access control layer 515 and/or enterprise access control module 918 ).

In the third stage, given the information graph 220 , the system can outline a process for information retrieval. One approach for this starts with first embedding the content of the text nodes (and/or other textual metadata associated with other modalities) (step 229 ) and storing the embeddings in a vector store 226 . Given a query 222 , the system can embed it (step 228 ) and find the most relevant text chunks or text nodes 230 associated with it. This would be the entry to the graph. At that point, the system can follow the out-going edges to other modality nodes. The classes associated with these modalities, (e.g., code, image, and table) may have a method that enables generation of relevant insights given the query. This method can be powered by different approaches, including multimodal models or other tools for understanding and querying the specific modalities. These insights, together with the text chunks and the query can then be combined in an aggregator 250 to a body of text or a prompt to be used for querying a model (e.g., multimodal model, large language model, etc.).

In one implementation, the functionality shown and described in FIG. 2 can be performed by a chunking module (e.g., chunking module 910 ) and/or embeddings generator module (e.g., embeddings generator module 912 ). For example, steps 204 - 213 may be performed by the chunking module, and the step 228 - 229 may be performed by the embeddings generator module. In some embodiments, the aggregator 250 comprises a portion of an orchestrator module (e.g., orchestrator module 904 ) and/or comprehension module (e.g., comprehension module 916 ).

FIG. 3 depicts a flowchart 300 of an example anti-hallucination and attribution process for enterprise generative artificial intelligence systems. Generally, the anti-hallucination and attribution process (e.g., performed by an anti-hallucination and attribution module 110 ) is a fully standalone process (and module) that can be used to validate and corroborate any response (e.g., a response from a generative artificial intelligence model or a human generated response), regardless of whether it comes with inline citations or not.

In step 302 , a response is received (e.g., from an enterprise generative artificial intelligence system, generative artificial intelligence model, human user, etc.). The response may be received by an anti-hallucination and attribution module (e.g., anti-hallucination and attribution module 110 ). In step 304 , the anti-hallucination and attribution module determines whether the response already includes citations (e.g., provided by the generative artificial intelligence model of the enterprise generative artificial intelligence system during the response generation process). If the anti-hallucination and attribution module determines the response is associated with citations (e.g., inline citations), the segmentation of the response can be extracted by segmenting based on the available citations (step 306 ). If the anti-hallucination and attribution module determines that citations are not present in the response, the anti-hallucination and attribution module performs generic segmentation (step 308 ). More specifically, the generic segmentation is generally conducted based on the number of tokens and based on the sentences in the response. More specifically, for each segment, the anti-hallucination and attribution module can fit in as many sentences in the segment until it hits a segment maximum token limit.

In step 310 , the anti-hallucination and attribution module assigns a set of sources based on a relevancy or similarity search. The anti-hallucination and attribution module can filter the source passages (or, sources) based on the resulting scores. The filtering may compare the resulting scores with one or more threshold values. For example, if a score is at or above the threshold value, then the associated source passages may be attributed to the segment. In another example, if a score is below the threshold value, then the associated source passage is not attributed to the segment. The similarity search can be an extension of a similarity score (e.g., “classical” cosine) and/or a function that can be over-written by an end-user.

The anti-hallucination and attribution module uses a mapping between the response segments and the sources with the relevancy score. The anti-hallucination and attribution module can also quantify/extract a credibility score associated with each source. The anti-hallucination and attribution module can then conduct a segmentation of the sources (step 312 ). More specifically, the anti-hallucination and attribution module can use the same segmentation logic used for segmenting the response. The anti-hallucination and attribution module computes the pairwise relevancy/similarity score among the response and source segments (step 314 ).

In step 316 , the anti-hallucination and attribution module further enhances this mapping with a corroboration score/label associated with each pair, within the four categories of contradiction, supporting, neutral and implication. These can then be used to further prune key and values in this mapping. The supporting label can be added to potentially cover for shortcomings associated with generic segmentation.

At this point, the anti-hallucination and attribution module has generated and quantified the relationship among the response and source segments. This is illustrated in FIG. 4 which shows all the different scores that define the relationships among the response and source segments. These can in turn be used to quantify a singular score associated with each response segment (e.g., quantifying its groundedness in sources) and a singular score associated with each source segment (e.g., in the context of the corresponding response segment). See step 318 . In some embodiments, this process can accommodate user-defined heuristics for computing some or all of these scores and for computing the source credibility score based on the available metadata of the source, either during inference or upon ingestion of the sources.

FIG. 4 depicts a diagram 400 of an example structure of response segments generated by an anti-hallucination and attribution method for enterprise generative artificial intelligence systems according to some embodiments. More specifically, FIG. 4 includes response segments 402 , sources, 404 , relevancy scores 406 for the sources 404 , credibility scores 408 , sources segments 410 , relevancy scores for sources segments 412 , and corroboration scores 414 .

FIG. 5 depicts a diagram of an example enterprise generative artificial intelligence system architecture and environment 500 according to some embodiments. In the example of FIG. 5 , the system architecture and environment 500 includes an enterprise generative artificial intelligence system 502 , enterprise systems 504 , external systems 506 , domain models 508 , a vector datastore 526 , an embeddings models datastore 524 , and an enterprise access control layer 515 . In some embodiments, the query comprises a natural language query received through a graphical user interface. In some embodiments, the one or more enterprise data sets include any of documents, document segments, and insights generated by the one or more artificial intelligence applications. In some embodiments, each of the relevance scores is associated with a respective portion of the one or

CLAIMS

Claims ( 20 )

What is claimed is:

1 . A method comprising:

receiving output from a generative artificial intelligence model processing a prompt; parsing the output from a generative model into chunks to be attributed to one or more source passages; retrieving the one or more source passages based on a similarity evaluation between the chunks and the one or more source passages; attributing at least a portion of the one or more source passages to the chunks based on a similarity threshold value; combining the chunks with the source passages attributed to those chunks; and generating a response to the prompt based on the combination, the response including the output with inline source identifiers that identify the attributed source passages.

2 . The method of claim 1 , wherein the generative artificial intelligence model employs a generative adversarial network, a variational autoencoder, an autoregressive model, or a recurrent neural network.

3 . The method of claim 1 , wherein the output includes sentences, and the chunks include one or more of the sentences.

4 . The method of claim 1 , wherein the similarity evaluation includes a similarity evaluation that calculates similarity scores associated with the chunks and the source passages.

5 . The method of claim 1 , wherein the response is generated by a generative artificial intelligence model.

6 . The method of claim 4 , further comprising filtering the one or more source passages based on the similarity scores.

7 . The method of claim 6 , further comprising:

generating an information graph for each data record in a set of data records including the one or more source passages, wherein the information graph describes relationships between source passages and one or more other classes, the other classes including source images, source tables, and source code.

8 . The method of claim 7 , wherein retrieving the one or more source passages is based on the information graph.

9 . A system comprising:

one or more processors; and memory storing instructions that, when executed by the one or more processors, cause the system to perform:

receiving output from a generative artificial intelligence model processing a prompt;

parsing the output from a generative model into chunks to be attributed to one or more source passages;

retrieving the one or more source passages based on a similarity evaluation between the chunks and the one or more source passages;

attributing at least a portion of the one or more source passages to the chunks based on a similarity threshold value;

combining the chunks with the source passages attributed to those chunks; and

generating a response to the prompt based on the combination, the response including the output with inline source identifiers that identify the attributed source passages.

10 . The system of claim 9 , wherein the generative artificial intelligence model employs a generative adversarial network, a variational autoencoder, an autoregressive model, or a recurrent neural network.

11 . The system of claim 9 , wherein the output includes sentences, and the chunks include one or more of the sentences.

12 . The system of claim 9 , wherein the similarity evaluation includes a similarity evaluation that calculates similarity scores associated with the chunks and the source passages.

13 . The system of claim 9 , wherein the response is generated by a generative artificial intelligence model.

14 . The system of claim 12 , wherein the instructions, when executed by the one or more processors, cause the system to perform filtering the one or more source passages based on the similarity scores.

15 . The system of claim 14 , wherein the instructions, when executed by the one or more processors, cause the system to perform generating an information graph for each data record in a set of data records including the one or more source passages, wherein the information graph describes relationships between source passages and one or more other classes, the other classes including source images, source tables, and source code.

16 . The system of claim 15 , wherein retrieving the one or more source passages is based on the information graph.

17 . A non-transitory computer readable medium comprising instructions that, when executed, cause one or more processors to perform:

receiving output from a generative artificial intelligence model processing a prompt; parsing the output from a generative model into chunks to be attributed to one or more source passages; retrieving the one or more source passages based on a similarity evaluation between the chunks and the one or more source passages; attributing at least a portion of the one or more source passages to the chunks based on a similarity threshold value; combining the chunks with the source passages attributed to those chunks; and generating a response to the prompt based on the combination, the response including the output with inline source identifiers that identify the attributed source passages.

18 . The non-transitory computer readable medium of claim 17 , wherein the generative artificial intelligence model employs a generative adversarial network, a variational autoencoder, an autoregressive model, or a recurrent neural network.

19 . The non-transitory computer readable medium of claim 17 , wherein the output includes sentences, and the chunks include one or more of the sentences.

20 . The non-transitory computer readable medium of claim 17 , wherein the similarity evaluation includes a similarity evaluation that calculates similarity scores associated with the chunks and the source passages.

US18/651,650

2023-05-01

2024-04-30

Enterprise generative artificial intelligence anti-hallucination and attribution architecture

Pending

US20240370709A1

( en )

Priority Applications (1)

Application Number

Priority Date

Filing Date

Title

US18/651,650

US20240370709A1

( en )

2023-05-01

2024-04-30

Enterprise generative artificial intelligence anti-hallucination and attribution architecture

Applications Claiming Priority (5)

Application Number

Priority Date

Filing Date

Title

US202363499485P

2023-05-01

2023-05-01

US18/542,536

US12111859B2

( en )

2022-12-16

2023-12-15

Enterprise generative artificial intelligence architecture

US18/542,583

US20240202539A1

( en )

2022-12-16

2023-12-15

Generative artificial intelligence crawling and chunking

US18/542,481

US12265570B2

( en )

2022-12-16

2023-12-15

Generative artificial intelligence enterprise search

US18/651,650

US20240370709A1

( en )

2023-05-01

2024-04-30

Enterprise generative artificial intelligence anti-hallucination and attribution architecture

Related Parent Applications (3)

Application Number

Title

Priority Date

Filing Date

US18/542,583

Continuation-In-Part

US20240202539A1

( en )

2022-12-16

2023-12-15

Generative artificial intelligence crawling and chunking

US18/542,536

Continuation-In-Part

US12111859B2

( en )

2022-12-16

2023-12-15

Enterprise generative artificial intelligence architecture

US18/542,481

Continuation-In-Part

US12265570B2

( en )

2022-12-16

2023-12-15

Generative artificial intelligence enterprise search

Publications (1)

Publication Number

Publication Date

US20240370709A1

true

US20240370709A1 ( en )

2024-11-07

Family

ID=93292772

Family Applications (1)

Application Number

Title

Priority Date

Filing Date

US18/651,650

Pending

US20240370709A1

( en )

2023-05-01

2024-04-30

Enterprise generative artificial intelligence anti-hallucination and attribution architecture

Country Status (1)

Country

Link

US

( 1 )

US20240370709A1

( en )

Cited By (19)

* Cited by examiner, † Cited by third party

Publication number

Priority date

Publication date

Assignee

Title

US20240378206A1

( en )

*

2023-05-08

2024-11-14

Ming Jack Po

System and method for answering questions requiring database query results in a large language model chat

US20250028977A1

( en )

*

2023-07-20

2025-01-23

Bank Of America Corporation

Artificial intelligence-based auditors of artificial intelligence

US20250103797A1

( en )

*

2023-09-26

2025-03-27

Dropbox, Inc.

Generating field objects for auto-populating fillable documents utilizing a large language model

CN119830331A

( en )

*

2025-01-17

2025-04-15

中国人民解放军国防科技大学

Large model attribution right confirming method, device and equipment for user right control

US20250190463A1

( en )

*

2023-10-23

2025-06-12

Riversound Solutions, LLC

Systems And Methods For Partial Information Retrieval Using Data Provenance Techniques

US20250265460A1

( en )

*

2024-02-15

2025-08-21

Reltio, Inc.

Flexibile entity resolution networks

US20250284716A1

( en )

*

2023-10-05

2025-09-11

Nasdaq, Inc.

Systems and methods of chained conversational prompt engineering for information retrieval

US20250299078A1

( en )

*

2024-03-20

2025-09-25

Bank Of America Corporation

Partial quantum mirror mode for artificial intelligence (ai) models

US20250335705A1

( en )

*

2024-04-25

2025-10-30

Robert Bosch Gmbh

System and method for knowledge-based audio-text modeling via automatic multimodal graph construction

US20250355894A1

( en )

*

2024-05-20

2025-11-20

View Systems, Inc.

Universal data representation for heterogeneous data

US12579159B2

( en )

*

2023-10-23

2026-03-17

Intuit Inc.

File extraction and vectorization for onboarding with LLM

US12596629B1

( en )

*

2025-06-04

2026-04-07

Future AGI Inc.

System for detecting errors in generative AI outputs

US20260148736A1

( en )

*

2024-11-22

2026-05-28

Google Llc

System For Generating Response To Multimodal Queries Using Multi-Step Reasoning

WO2026122935A1

( en )

*

2024-12-06

2026-06-11

Google Llc

Managing and using user-specific conditioning data

WO2026128206A1

( en )

*

2024-12-13

2026-06-18

Qualcomm Incorporated

Enforcing self-consistency in language model outputs

US20260211997A1

( en )

*

2025-01-23

2026-07-23

International Business Machines Corporation

Secure Runtime Enforcement of Agentic Workflows based on Tool Invocation Intent

US20260211889A1

( en )

*

2025-01-23

2026-07-23

Google Llc

System for controlling response reliability

US12694227B1

( en )

*

2025-08-05

2026-07-28

Zeev Barkan

System of correcting LLM hallucinations

US12711308B2

( en )

*

2024-04-25

2026-08-18

Robert Bosch Gmbh

System and method for knowledge-based audio-text modeling via automatic multimodal graph construction

Citations (8)

* Cited by examiner, † Cited by third party

Publication number

Priority date

Publication date

Assignee

Title

US20070055656A1

( en )

*

2005-08-01

2007-03-08

Semscript Ltd.

Knowledge repository

US20110307435A1

( en )

*

2010-05-14

2011-12-15

True Knowledge Ltd

Extracting structured knowledge from unstructured text

US10943072B1

( en )

*

2019-11-27

2021-03-09

ConverSight.ai, Inc.

Contextual and intent based natural language processing system and method

US20220237368A1

( en )

*

2021-01-22

2022-07-28

Bao Tran

Systems and methods for machine content generation

US20230061906A1

( en )

*

2021-08-09

2023-03-02

Samsung Electronics Co., Ltd.

Dynamic question generation for information-gathering

US20230274094A1

( en )

*

2021-08-24

2023-08-31

Unlikely Artificial Intelligence Limited

Computer implemented methods for the automated analysis or use of data, including use of a large language model

US20240112394A1

( en )

*

2022-09-29

2024-04-04

Lifecast Incorporated

AI Methods for Transforming a Text Prompt into an Immersive Volumetric Photo or Video

US20240320867A1

( en )

*

2023-03-20

2024-09-26

Sony Interactive Entertainment Inc.

Iterative Image Generation From Text

2024

2024-04-30

US

US18/651,650

patent/US20240370709A1/en

active

Pending

Patent Citations (9)

* Cited by examiner, † Cited by third party

Publication number

Priority date

Publication date

Assignee

Title

US20070055656A1

( en )

*

2005-08-01

2007-03-08

Semscript Ltd.

Knowledge repository

US8666928B2

( en )

*

2005-08-01

2014-03-04

Evi Technologies Limited

Knowledge repository

US20110307435A1

( en )

*

2010-05-14

2011-12-15

True Knowledge Ltd

Extracting structured knowledge from unstructured text

US10943072B1

( en )

*

2019-11-27

2021-03-09

ConverSight.ai, Inc.

Contextual and intent based natural language processing system and method

US20220237368A1

( en )

*

2021-01-22

2022-07-28

Bao Tran

Systems and methods for machine content generation

US20230061906A1

( en )

*

2021-08-09

2023-03-02

Samsung Electronics Co., Ltd.

Dynamic question generation for information-gathering

US20230274094A1

( en )

*

2021-08-24

2023-08-31

Unlikely Artificial Intelligence Limited

Computer implemented methods for the automated analysis or use of data, including use of a large language model

US20240112394A1

( en )

*

2022-09-29

2024-04-04

Lifecast Incorporated

AI Methods for Transforming a Text Prompt into an Immersive Volumetric Photo or Video

US20240320867A1

( en )

*

2023-03-20

2024-09-26

Sony Interactive Entertainment Inc.

Iterative Image Generation From Text

Non-Patent Citations (2)

* Cited by examiner, † Cited by third party

Title

Baolin Peng, Check Your Facts and Try Again: Improving Large Language Models with External Knowledge and Automated Feedback, March 8, 2023, Microsoft Research, Columbia University (Year: 2023), Pages 1 - 15

*

Liu et al., " A Token-level Reference-free Hallucination Detection Benchmark for Free-form Text Generation ", (Year: 2022)

*

Cited By (22)

* Cited by examiner, † Cited by third party

Publication number

Priority date

Publication date

Assignee

Title

US20240378206A1

( en )

*

2023-05-08

2024-11-14

Ming Jack Po

System and method for answering questions requiring database query results in a large language model chat

US20250028977A1

( en )

*

2023-07-20

2025-01-23

Bank Of America Corporation

Artificial intelligence-based auditors of artificial intelligence

US12626152B2

( en )

*

2023-07-20

2026-05-12

Bank Of America Corporation

Artificial intelligence-based auditors of artificial intelligence

US12462096B2

( en )

*

2023-09-26

2025-11-04

Dropbox, Inc.

Generating field objects for auto-populating fillable documents utilizing a large language model

US20250103797A1

( en )

*

2023-09-26

2025-03-27

Dropbox, Inc.

Generating field objects for auto-populating fillable documents utilizing a large language model

US20250284716A1

( en )

*

2023-10-05

2025-09-11

Nasdaq, Inc.

Systems and methods of chained conversational prompt engineering for information retrieval

US20250190463A1

( en )

*

2023-10-23

2025-06-12

Riversound Solutions, LLC

Systems And Methods For Partial Information Retrieval Using Data Provenance Techniques

US12579159B2

( en )

*

2023-10-23

2026-03-17

Intuit Inc.

File extraction and vectorization for onboarding with LLM

US12619876B2

( en )

*

2024-02-15

2026-05-05

Reltio, Inc.

Flexible entity resolution networks

US20250265460A1

( en )

*

2024-02-15

2025-08-21

Reltio, Inc.

Flexibile entity resolution networks

US20250299078A1

( en )

*

2024-03-20

2025-09-25

Bank Of America Corporation

Partial quantum mirror mode for artificial intelligence (ai) models

US20250335705A1

( en )

*

2024-04-25

2025-10-30

Robert Bosch Gmbh

System and method for knowledge-based audio-text modeling via automatic multimodal graph construction

US12711308B2

( en )

*

2024-04-25

2026-08-18

Robert Bosch Gmbh

System and method for knowledge-based audio-text modeling via automatic multimodal graph construction

US20250355894A1

( en )

*

2024-05-20

2025-11-20

View Systems, Inc.

Universal data representation for heterogeneous data

US20260148736A1

( en )

*

2024-11-22

2026-05-28

Google Llc

System For Generating Response To Multimodal Queries Using Multi-Step Reasoning

WO2026122935A1

( en )

*

2024-12-06

2026-06-11

Google Llc

Managing and using user-specific conditioning data

WO2026128206A1

( en )

*

2024-12-13

2026-06-18

Qualcomm Incorporated

Enforcing self-consistency in language model outputs

CN119830331A

( en )

*

2025-01-17

2025-04-15

中国人民解放军国防科技大学

Large model attribution right confirming method, device and equipment for user right control

US20260211997A1

( en )

*

2025-01-23

2026-07-23

International Business Machines Corporation

Secure Runtime Enforcement of Agentic Workflows based on Tool Invocation Intent

US20260211889A1

( en )

*

2025-01-23

2026-07-23

Google Llc

System for controlling response reliability

US12596629B1

( en )

*

2025-06-04

2026-04-07

Future AGI Inc.

System for detecting errors in generative AI outputs

US12694227B1

( en )

*

2025-08-05

2026-07-28

Zeev Barkan

System of correcting LLM hallucinations

Similar Documents

Publication

Publication Date

Title

US12111859B2

( en )

2024-10-08

Enterprise generative artificial intelligence architecture

US11640494B1

( en )

2023-05-02

Systems and methods for construction, maintenance, and improvement of knowledge representations

US11727000B2

( en )

2023-08-15

System and method for reducing resource usage in a data retrieval process

US9535902B1

( en )

2017-01-03

Systems and methods for entity resolution using attributes from structured and unstructured data

US10223417B1

( en )

2019-03-05

System and method for reducing query-related resource usage in a data retrieval process

US20230075655A1

( en )

2023-03-09

Systems and methods for context-independent database search paths

US20240354584A1

( en )

2024-10-24

Language model-based tabular data object extraction and visualization

AU2024205713A1

( en )

2024-09-12

Multi-source-type interoperability and/or information retrieval optimization

KR20250054242A

( en )

2025-04-23

A system that provides questions and answers about space technology output data using an artificial intelligence model

WO2025170605A1

( en )

2025-08-14

Enterprise generative artificial intelligence anti-hallucination and attribution architecture

Betgeri et al.

2025

Automating SQL Databases using Artificial Intelligence and Natural Language Processing

Legal Events

Date

Code

Title

Description

2024-11-04

STPP

Information on status: patent application and granting procedure in general

Free format text : RESPONSE TO NON-FINAL OFFICE ACTION ENTERED AND FORWARDED TO EXAMINER

2024-11-29

STPP

Information on status: patent application and granting procedure in general

Free format text : NON FINAL ACTION MAILED

2025-05-19

AS

Assignment

Owner name : C3.AI, INC., CALIFORNIA

Free format text : ASSIGNMENT OF ASSIGNORS INTEREST;ASSIGNORS:SIEBEL, THOMAS M.;PAKAZAD, SINA KHOSHFETRAT;JUBAN, ROMAIN;AND OTHERS;SIGNING DATES FROM 20240508 TO 20250423;REEL/FRAME:071160/0222

2025-06-30

STPP

Information on status: patent application and granting procedure in general

Free format text : NON FINAL ACTION MAILED

2025-10-01

STPP

Information on status: patent application and granting procedure in general

Free format text : RESPONSE TO NON-FINAL OFFICE ACTION ENTERED AND FORWARDED TO EXAMINER

2025-10-10

STPP

Information on status: patent application and granting procedure in general

Free format text : RESPONSE TO NON-FINAL OFFICE ACTION ENTERED AND FORWARDED TO EXAMINER

2026-06-05

STPP

Information on status: patent application and granting procedure in general

Free format text : FINAL REJECTION COUNTED, NOT YET MAILED

2026-06-10

STPP

Information on status: patent application and granting procedure in general

Free format text : FINAL REJECTION MAILED

Related documents

• Generative AI and the Limits of Work Product Doctrine#1174756
• Generative AI May Reinforce Social Biases in Software Engineering Education#1076140
• AI, Decomputing and the Interregnum#217790
• Inexcuseable? Building an Artificial Intelligence Microexpert on Contract Excuse Doctrine#228780
• Growing the image: Generative AI and the medium of gardening#87689