๊ฐ€์ž… ํ›„ ์ดˆ๋Œ€ ๋งํฌ๋ฅผ ๊ณต์œ ํ•˜๋ฉด ๋™์˜์ƒ ์žฌ์ƒ ๋ฐ ์ดˆ๋Œ€ ๋ณด์ƒ์„ ๋ฐ›์„ ์ˆ˜ ์žˆ์Šต๋‹ˆ๋‹ค.

Deli Chen
@victor207755822
Deep Learning Researcher @deepseek_ai | #AGIforEveryone# Prev. BS and MS @PKU1898 | | All opinions are my own. | INTP-T | ไบบๅฟƒๆƒŸๅฑ๏ผŒ้“ๅฟƒๆƒŸๅพฎ
๊ฐ€์ž… December 2023
181 ํŒ”๋กœ์ž‰ ์ค‘    34.1K ํŒฌ
๐Ÿงต Deli AutoResearch SKILL is now officially open source! ๐ŸŽ‰ Alongside it, weโ€™re dropping our 4th survey paper โ€” this time on Self-play. Inspired by AlphaZero, we got a powerful insight: prior knowledge doesnโ€™t always lift the ceiling. Models can discover more globally optimal solutions just by playing against themselves. The biggest change in this paper? For the first time, the AutoResearch Agent autonomously planned GPU experiments โ€” and submitted actual RL runs on the DeepSeek 285B model. The entire RL pipeline โ€” experiment design, code writing, running, debugging, and conclusion summarization โ€” was 100% automated, with zero human intervention from me. This was incredibly difficult, but an incredibly important step. GRPO is the tool being called by the AutoResearch Agent here. We see this as the beginning of our Continual Learning research journey. ๐Ÿš€ As always, this is my personal research project, unaffiliated with any organization. All views are my own. #AI# #ReinforcementLearning# #SelfPlay# #OpenSource# #AutoML# #ContinualLearning# #DeepSeek#
๋” ๋ณด๊ธฐ