US 2026/0300784 A1 · ADAPTIVE PLANNING USING LLMS

Abstract

Systems, methods, and other embodiments associated with adaptive planning using LLMs are described. In one embodiment, a method includes receiving a digital representation of a question expressed in natural language for resolution by a large language model. The method includes acquiring digital information that is relevant to answering the question and absent from a digital context associated with the question. The method includes determining that a gap exists in the acquired information based on assessment by the large language model of whether the acquired information is sufficient to answer the question. The method generates a sub-question based on detecting divergent dimensions among the embeddings, wherein the sub-question is formulated to request information that fills the gap in the acquired information. And the method includes recursively executing the computer-implemented method with the sub-question in place of the question until the gap in the acquired information is eliminated.

Claims20

Dependent claims

A computer-implemented method, comprising:

intercept, by a computing system, a digital representation of a question expressed in natural language presented for resolution by a large language model;
acquiring, by the computing system, digital information that is relevant to answering the question and absent from a digital context associated with the natural language question;
determining, by the computing system, that a gap exists in the acquired information based on an information sufficiency response by the large language model to an information sufficiency query to assess whether the acquired information is sufficient to answer the question;
generating, by the computing system, a sub-question in natural language that is formulated to request information that fills the gap in the acquired information;
and
recursively executing, by the computing system, the computer-implemented method with the sub-question in place of the question until the gap in the acquired information is eliminated, thereby improving accuracy and efficiency of reasoning by the large language model by adaptively identifying and filling gaps in the acquired information.

3 more elements

A computing system, comprising:

at least one processor connected to at least one memory;
one or more non-transitory computer readable media having instructions stored thereon that when executed by at least the processor cause the computing system to execute steps of a computer-implemented method, comprising:
intercept a digital representation of a question expressed in natural language presented for resolution by a large language model;
acquire digital information that is relevant to answering the question and absent from a digital context associated with the natural language question;
determine that a gap exists in the acquired information based on an information sufficiency response by the large language model to an information sufficiency query to assess whether the acquired information is sufficient to answer the question;
generate a sub-question in natural language that is formulated to request information that fills the gap in the acquired information;
and
recursively execute the computer-implemented method with the sub-question in place of the question until the gap in the acquired information is eliminated.

5 more elements

One or more non-transitory computer readable media having instructions stored thereon that when executed by at least a processor of a computing system cause the computing system to execute steps of a computer-implemented method, comprising:

intercept a digital representation of a question expressed in natural language presented for resolution by a large language model;
acquire digital information that is relevant to answering the question and absent from a digital context associated with the natural language question;
determine that a gap exists in the acquired information based on an information sufficiency response by the large language model to an information sufficiency query to assess whether the acquired information is sufficient to answer the question;
generate a sub-question in natural language that is formulated to request information that fills the gap in the acquired information;
and
recursively execute the computer-implemented method with the sub-question in place of the question until the gap in the acquired information is eliminated.

3 more elements

Dependencies are inferred from the claim text.

Adaptive planning using llms

Life of this patent

Adaptive planning using llms

Life of this patent

Abstract

Claims20