Systems for controllable summarization of content
A method of generating summaries of content items using one or more large language models (LLMs) is disclosed. A first content item is identified. The first content item includes a set of sub-content items. A level of abstraction is determined for the content item. A prompt is automatically engineered for providing to the one or more LLMs. The prompt includes a reference to the first content item and the level of the abstraction for the first content item. A response to the prompt is received from the LLM. The response includes a second content item. The second content item includes a representation of the first content item that is generated by the LLM. The representation omits or simplifies one or more of the set of sub-content items based on the level of abstraction. The representation is used to control an output that is communicated to a target device.
1. A system comprising:
one or more computer-implemented subsystems, the one or more computer-implemented subsystems configured to perform operations, the operations comprising:
identifying a first content item, the first content item including a set of sub-content items;
determining a level of abstraction for the first content item;
automatically engineering a prompt for providing to a large language model (LLM), the prompt including a reference to the first content item and the level of the abstraction for the first content item;
receiving a response to the prompt from the LLM, the response including a second content item, the second content item including a representation of the first content item that is generated by the LLM, the representation omitting or simplifying one or more of the set of sub-content items based on the level of abstraction; and
using the representation to control an output that is communicated to a target device.
2. The system of claim 1 , wherein the level of the abstraction is specified as a percentage of a measurement pertaining to the first content item that is to be used by the LLM to limit a corresponding measurement of the representation.
3. The system of claim 2 , wherein the measurement pertaining to the first content item relates to a length or size of the first content item and the corresponding measurement of the representation relates to a length or size of the representation.
4. The system of claim 1 , wherein the determining of the level of the abstraction is based on an input received from a user via a graphical user interface.
5. The system of claim 1 , the operations further comprising, based on the level of abstraction transgressing a configurable threshold value, further engineering the prompt to request that the LLM make one or more inferences based on the first content item and to generate information pertaining to the one or more inferences for including in the representation.
6. The system of claim 5 , the operations further comprising, further engineering the prompt to request that the LLM use one or more external data sources to generate the information pertaining the one or more inferences.
7. The system of claim 1 , wherein the using of the representation includes engineering an additional prompt to the LLM, the additional prompt requesting the LLM to generate an appropriate conversational response to the representation based on a context associated with the representation and an entity represented by the system.
8. The system of claim 7 , wherein the entity is a sales representative and the context includes information pertaining to a sales call being handled in real time by the sales representative.
9. The system of claim 8 , wherein the sub-content items include one or more statements made by a customer and one or more statement made by the sales representative.
10. The system of claim 9 , wherein the sales representative is an intelligent agent.
11. A method comprising:
identifying a first content item, the first content item including a set of sub-content items;
determining a level of abstraction for the first content item;
automatically engineering a prompt for providing to a large language model (LLM), the prompt including a reference to the first content item and the level of the abstraction for the first content item;
receiving a response to the prompt from the LLM, the response including a second content item, the second content item including a representation of the first content item that is generated by the LLM, the representation omitting or simplifying one or more of the set of sub-content items based on the level of abstraction; and
using the representation to control an output that is communicated to a target device.
12. The method of claim 11 , further comprising, based on the level of abstraction transgressing a configurable threshold value, further engineering the prompt to request that the LLM make one or more inferences based on the first content item and to generate information pertaining to the one or more inferences for including in the representation.
13. The method of claim 12 , further comprising further engineering the prompt to request that the LLM use one or more external data sources to generate the information pertaining the one or more inferences.
14. The method of claim 11 , wherein the using of the representation includes engineering an additional prompt to the LLM, the additional prompt requesting the LLM to generate an appropriate conversational response to the representation based on a context associated with the representation and an entity represented by a system.
15. The method of claim 14 , wherein the entity is a sales representative and the context includes information pertaining to a sales call being handled in real time by the sales representative.
16. A non-transitory computer-readable storage medium storing a set of instructions that, when executed by one or more computer processors, causes the one or more computer processors to perform operations, the operations comprising:
identifying a first content item, the first content item including a set of sub-content items;
determining a level of abstraction for the first content item;
automatically engineering a prompt for providing to a large language model (LLM), the prompt including a reference to the first content item and the level of the abstraction for the first content item;
receiving a response to the prompt from the LLM, the response including a second content item, the second content item including a representation of the first content item that is generated by the LLM, the representation omitting or simplifying one or more of the set of sub-content items based on the level of abstraction; and
using the representation to control an output that is communicated to a target device.
17. The non-transitory computer-readable storage medium of claim 16 , the operations further comprising, based on the level of abstraction transgressing a configurable threshold value, further engineering the prompt to request that the LLM make one or more inferences based on the first content item and to generate information pertaining to the one or more inferences for including in the representation.
18. The non-transitory computer-readable storage medium of claim 17 , the operations further comprising further engineering the prompt to request that the LLM use one or more external data sources to generate the information pertaining the one or more inferences.
19. The non-transitory computer-readable storage medium of claim 16 , wherein the using of the representation includes engineering an additional prompt to the LLM, the additional prompt requesting the LLM to generate an appropriate conversational response to the representation based on a context associated with the representation and an entity represented by a system.
20. The non-transitory computer-readable storage medium of claim 19 , wherein the entity is a sales representative and the context includes information pertaining to a sales call being handled in real time by the sales representative.