Abstract: A computer-implemented method for generating an output token sequence from an input token sequence. The method combines a look ahead tree search, such as a Monte Carlo tree search, with a sequence-to-sequence neural network system. The sequence-to-sequence neural network system has a policy output defining a next token probability distribution, and may include a value neural network providing a value output to evaluate a sequence. An initial partial output sequence is extended using the look ahead tree search guided by the policy output and, in implementations, the value output, of the sequence-to-sequence neural network system until a complete output sequence is obtained.

Abstract: In one embodiment, a method for broadcast state agnostic timeline metadata presentation on a companion device is implemented on a computing device and includes: receiving timeline metadata associated with a video content item in a pseudo-live mode, where the timeline metadata is received from a timeline metadata provision system according to a timeline for an original broadcast of the video content item, and presenting the timeline metadata as it is received from said timeline metadata provision system, where the presenting is in parallel to an original or non-original broadcast of the video content item playing on an associated viewing device.

Abstract: There are instances of a content item having synchronization information, one being an instance of a reference content item, another being a candidate content item. A synchronizing processor determines a synchronization point between the reference content item and the candidate content item, on the basis of matching the synchronization information of the reference content item and the candidate content item. A timeline metadata transmitter transmits stored timeline metadata to a device on which the reference content item is playing, the transmission of the stored timeline metadata beginning from the synchronization point. The stored timeline metadata includes timeline metadata that has been previously aggregated from earlier broadcasts of instances of the reference content item. Related systems, apparatus, and methods are also described.

Abstract: In one embodiment, a first content item having synchronization information, the first content item being a first instance of a reference content item, a candidate content item, which is a second instance of the reference content item, having synchronization information, a synchronizing processor which determines a synchronization point between the first content item and the candidate content item, on the basis of matching the synchronization information of the first content item and the candidate content item, a timeline metadata transmitter, which transmits stored timeline metadata to a device on which the first content item is playing, the transmission of the stored timeline metadata beginning from the synchronization point, wherein the stored timeline metadata includes timeline metadata that has been previously aggregated from earlier broadcasts of instances of the reference content item. Related systems, apparatus, and methods are also described.

Abstract: In one embodiment, a method for broadcast state agnostic timeline metadata presentation on a companion device is implemented on a computing device and includes: receiving timeline metadata associated with a video content item in a pseudo-live mode, where the timeline metadata is received from a timeline metadata provision system according to a timeline for an original broadcast of the video content item, and presenting the timeline metadata as it is received from said timeline metadata provision system, where the presenting is in parallel to an original or non-original broadcast of the video content item playing on an associated viewing device.

Abstract: Delivering an audio video asset in a way that makes best use of different delivery mechanisms available.

Abstract: Delivering an audio video asset in a way that makes best use of different delivery mechanisms available.

Abstract: A method of delivering an audio video asset is described.

Abstract: A method of delivering an audio video asset is described.