전체 글 썸네일형 리스트형 AlphaGeometry： Solving olympiad geometry without human demonstrations https://www.nature.com/articles/s41586-023-06747-5 문제를 입력받아 다음과 같은 도메인으로 1차 처리 이후에 이를 바탕으로 이게 되는지 안되는지 계속 tree search후, 정답 조건식들을 만족하게 되면 정답을 맞추는 방식 trial and error를 계속 사용하되 그것이 llm을 사용하게 하는 것으로 오히려 강화학습에 가깝지 않나 싶음 결국 무한의 도메인을 더 좁은 도메인으로 좁히는 방식으로 푸는 것 하지만 무한한 trial and error도 풀지 못한 5개가 존재하는데, 이에 대해서 알려주었다면 더 좋겠음 좋은 발전이지만 아직 논리는 이해한다고 보긴 어려워보임 더보기 Mixtral of Experts https://arxiv.org/abs/2401.04088 Mixtral of Experts We introduce Mixtral 8x7B, a Sparse Mixture of Experts (SMoE) language model. Mixtral has the same architecture as Mistral 7B, with the difference that each layer is composed of 8 feedforward blocks (i.e. experts). For every token, at each layer, a router arxiv.org 간단하게 표현하면 x개의 모델이 존재하며 앞에 FFN을 추가해서 그중 높은 점수 2개를 결합하여 사용하는 방식 Routing에 대한 인공지능을 .. 더보기 Voyager: An Open-Ended Embodied Agent with Large Language Models https://arxiv.org/abs/2305.16291?utm_source=substack&utm_medium=email Voyager: An Open-Ended Embodied Agent with Large Language Models We introduce Voyager, the first LLM-powered embodied lifelong learning agent in Minecraft that continuously explores the world, acquires diverse skills, and makes novel discoveries without human intervention. Voyager consists of three key components: 1) an arxiv... 더보기 Generative Agents: Interactive Simulacra of Human Behavior https://arxiv.org/abs/2304.03442 Generative Agents: Interactive Simulacra of Human Behavior Believable proxies of human behavior can empower interactive applications ranging from immersive environments to rehearsal spaces for interpersonal communication to prototyping tools. In this paper, we introduce generative agents--computational software ag arxiv.org LLM 자가 학습 더보기 재미있는 github https://github.com/KU-BIG/KUBIG_2023_FALL GitHub - KU-BIG/KUBIG_2023_FALL: Repository for codes from study sessions and projects during KUBIG 2023 FALL (학회장 염윤 Repository for codes from study sessions and projects during KUBIG 2023 FALL (학회장 염윤석 장수혁) - GitHub - KU-BIG/KUBIG_2023_FALL: Repository for codes from study sessions and projects during KUBIG 2023... github.com 더보기 LLaMA Pro： Progressive LLaMA with Block Expansion (Paper Explained) https://arxiv.org/html/2401.02415v1 LLaMA Pro: Progressive LLaMA with Block Expansion HTML conversions sometimes display errors due to content that did not convert correctly from the source. This paper uses the following packages that are not yet supported by the HTML conversion tool. Feedback on these issues are not necessary; they are kno arxiv.org 짧은 요약 LLaMA Pro는 기존 LLaMA 언어 모델을 기반으로 새로운 블록 .. 더보기 AI Learns Stairs (deep reinforcement learning) https://www.youtube.com/watch?v=xk8wHY1AFpI 잼나넹넹 더보기 InstantID https://instantid.github.io/ InstantID There has been significant progress in personalized image synthesis with methods such as Textual Inversion, DreamBooth, and LoRA. Yet, their real-world applicability is hindered by high storage demands, lengthy fine-tuning processes, and the need for multi instantid.github.io https://github.com/InstantID/InstantID GitHub - InstantID/InstantID: InstantID : Z.. 더보기 이전 1 ··· 42 43 44 45 46 47 48 ··· 62 다음
