Programming Problems Submitted for Evaluation of LLMs GPT-3.5 and Gemini Pro 1.0
收藏资源简介：
Problems extracted from platforms LeetCode and BeeCrowd for evaluation of LLMs GPT3.5 and Gemini Pro 1.0. The data from the plataforms has the following columns and values: LeetCode Data BeeCrowd Data Column Doc Column Doc problem_level easy | medium | hard problem_level <1...10> problem_link <LeetCode link for the problem> problem_link <BeeCrowd link for the problem> prompt <text submitted to LLM> prompt <text submitted to LLM> response_code <code provided by the LLM> response_code <code provided by the LLM> response_evaluation True | False response_evaluation True | False execution_time_ms <time> execution_time_ms <time> memory_usage_mb <memory> error_generated Wrong Answer | Time Limit Exceeded | Memory Limit Exceeded.... error_generated Wrong Answer | Time Limit Exceeded | Memory Limit Exceeded.... attempts_number 1 | 2 | 3 attempts_number 1 | 2 | 3 author <author's name> contains_image True | False source <origin institution> related_topic_1 <topic> origin_country <origin country> related_topic_2 <topic> contains_image True | False related_topic_3 <topic> related_topic_1 <topic> related_topic_4 <topic> related_topic_5 <topic>
