Abstract: The present disclosure relates to medicament injection devices. An injection device includes: a movable dosage programming component comprising a rotary encoder system having a predefined angular periodicity, a sensor arrangement including a first optical sensor configured to detect movement of the movable dosage programming component relative to the sensor arrangement during dosing of a medicament, wherein the first optical sensor is configured to operate in a strobe-sampling mode at a first frequency, a second optical sensor configured to detect movement of the rotary encoder system relative to the second optical sensor wherein the second optical sensor is configured to operate in a strobe-sampling mode at a second frequency lower than the first frequency, and a processor arrangement configured to, based on the detected movement, determine a medicament dosage administered by the injection device.

Abstract: Embodiments use a multi-modal approach to generate software programs that match a solution program description. The solution program description may include natural language, input-output examples, partial source code, desired operators, or other hints. Some embodiments use optimized prompts to a pre-trained language model to obtain initial candidate programs. Maximal program components are extracted and then recombined variously using component-based synthesis. Beam search reduces a solution program search space by discarding some candidates from a given synthesis iteration. Relevance metrics, string similarity metrics, operator frequency distributions, token rareness scores, and other optimizations may be employed. By virtue of optimizations and the multi-modal approach, a solution program may be obtained after fewer iterations than by use of a language model alone.

Abstract: Embodiments use a multi-modal approach to generate software programs that match a solution program description. The solution program description may include natural language, input-output examples, partial source code, desired operators, or other hints. Some embodiments use optimized prompts to a pre-trained language model to obtain initial candidate programs. Maximal program components are extracted and then recombined variously using component-based synthesis. Beam search reduces a solution program search space by discarding some candidates from a given synthesis iteration. Relevance metrics, string similarity metrics, operator frequency distributions, token rareness scores, and other optimizations may be employed. By virtue of optimizations and the multi-modal approach, a solution program may be obtained after fewer iterations than by use of a language model alone.