Google has developed the logic and reasoning skills of Bard, the chatbot based on the LaMDA major language model. The improvement will help Bard perform better on questions about math, logic, and programming.
The algorithm will now run and test code in the background before answering math tasks and coding questions. The result obtained will be used to improve the final answer. Bard has stepped up in answering complex questions with the latest update.
Calculation accuracy increased by 30%
Google’s in-house tests showed that Bard handled computationally demanding tasks better than before. The accuracy of the chatbot’s answers to computation-based word and math problems increased by 30%.
Google Bard was announced in early February. The chatbot, which was made available to everyone in March, continues to gain new abilities day by day.
38770
