Abstract: Embodiments are disclosed for long document question answering using large language models. The method may include receiving a question for a document. A representation of the document may then be obtained. A large language model (LLM) is used to identify one or more sections of the document that are relevant to the question using the representation of the document. A document question answering model determines an answer to the question using the one or more sections of the document.

Abstract: The disclosed technology includes an example system that has a request throttling manager that is configured to receive a first file data request, queue the first file data request in a first request queue, and process the first file data request based on the first token bucket. The first token bucket includes a sufficient first quantity of first tokens to process the first file data request. The system further includes a storage manager configured to access one or more storage nodes of a plurality of storage nodes of a distributed storage system in response to the first file data request.

Abstract: Systems and methods for zero-footprint and safe execution of quantum computing programs are disclosed. According to one embodiment, in an electronic device comprising at least one computer processor, a method for cloud-based execution of quantum-computing programs may include: (1) receiving, from a user interface on a client device, a serialized file comprising a domain, an application, and an algorithm; (2) receiving, from the user interface, problem data and an identification of a quantum computing backend for executing the problem data; (3) instantiating a quantum program for execution and communicating the quantum program and the problem data to the quantum computing backend for execution; (4) receiving, from the quantum computing backend, an output of the execution; and (5) communicating the output to the user interface on the client device.

Abstract: Systems and methods for zero-footprint and safe execution of quantum computing programs are disclosed. According to one embodiment, in an electronic device comprising at least one computer processor, a method for cloud-based execution of quantum-computing programs may include: (1) receiving, from a user interface on a client device, a serialized file comprising a domain, an application, and an algorithm; (2) receiving, from the user interface, problem data and an identification of a quantum computing backend for executing the problem data; (3) instantiating a quantum program for execution and communicating the quantum program and the problem data to the quantum computing backend for execution; (4) receiving, from the quantum computing backend, an output of the execution; and (5) communicating the output to the user interface on the client device.

Abstract: The disclosed technology includes an example system that has a request throttling manager that is configured to receive a first file data request, queue the first file data request in a first request queue, and process the first file data request based on the first token bucket. The first token bucket includes a sufficient first quantity of first tokens to process the first file data request. The system further includes a storage manager configured to access one or more storage nodes of a plurality of storage nodes of a distributed storage system in response to the first file data request.

Abstract: An efficient mechanism for wire-tapping network traffic is disclosed. In one embodiment of the invention, a primary forwarding lookup process and a secondary forwarding lookup process are performed in parallel and independently of each other. The primary forwarding lookup process determines the output interface to which the packet is to be routed regardless of whether the packet is to be intercepted. The secondary forwarding lookup process determines whether the packet is to be intercepted and also determines the output interface to which a copy of the packet is to be routed. Because the lookup processes are performed independently and in parallel, normal packet forwarding can be performed at line rate or near line rate while the packets are intercepted.

Abstract: Performing cascading lookups at a network node involves using first header information from a packet to search a first set of search criteria for a match, identifying a tag associated with the match, and using second header information from the packet and the tag to search a second set of search criteria for another match.

Abstract: An efficient mechanism for wire-tapping network traffic is disclosed. In one embodiment of the invention, a primary forwarding lookup process and a secondary forwarding lookup process are performed in parallel and independently of each other. The primary forwarding lookup process determines the output interface to which the packet is to be routed regardless of whether the packet is to be intercepted. The secondary forwarding lookup process determines whether the packet is to be intercepted and also determines the output interface to which a copy of the packet is to be routed. Because the lookup processes are performed independently and in parallel, normal packet forwarding can be performed at line rate or near line rate while the packets are intercepted.

Abstract: Performing cascading lookups at a network node involves using first header information from a packet to search a first set of search criteria for a match, identifying a tag associated with the match, and using second header information from the packet and the tag to search a second set of search criteria for another match.