使用langchain进行任务规划，构建子任务的会话场景资源，通过MCTS任务执行器，来让每个子任务通过在上下文中资源，通过自身反思探索来获取自身对问题的最优答案；这种方式依赖模型的对齐偏好，我们在每种偏好上设计了一个工程框架，来完成自我对不同答案的奖励进行采样策略
-
Updated
Mar 20, 2025 - Jupyter Notebook
#
mcts-agents
Here are 13 public repositories matching this topic...
An AI agent developed to play Ms. Pac-Man by adopting a strategy formed by MCTS and a FSM.
-
Updated
Jul 17, 2021 - C#
MiniMax with Alpha-Beta pruning and Monte-Carlo Tree Search implementations for the board game Hex
-
Updated
Jun 8, 2021 - Python
This is work-in-progress (WIP) refactored implementation of "Retreival-guided Reinforcement Learning for Boolean Circuit Minimization" work published in ICLR 2024.
reinforcement-learning deep-learning mcts logic-synthesis electronic-design-automation boolean-optimization electronic-design mcts-agents iclr2024 abc-rl retrieval-guided-rl boolean-circuit-minimization ml-for-logic-synthesis ml-for-chip-design
-
Updated
May 10, 2024 - Verilog
A Monte-Carlo Tree Search mathod that enables two agents interact and work together in the game of Pacman Capture the Flag.
-
Updated
Sep 19, 2023 - Python
Monte Carlo Tree Search
-
Updated
Jan 28, 2025
Lightweight, extensible, and fair multi- DNN manager for heterogeneous embedded devices.
deep-neural-networks deep-learning scheduler python3 pytorch embedded-systems performance-prediction workload-automation edge-inference mcts-agents heterogeneous-architectures concurrent-execution throughput-estimation
-
Updated
Jul 12, 2023 - Python
AI implementation using monte carlo tree search (MCTS) for the Game of Amazons
-
Updated
Oct 13, 2023 - C++
An AI agent for the card game Coup that uses ISMCTS.
-
Updated
Aug 28, 2023 - Python
Using reinforcement learning to play games.
game python games reinforcement-learning ai python3 artificial-intelligence mcts reinforcement-learning-algorithms artificial-neural-networks python-3 monte-carlo-tree-search reinforcement-learning-agent mcts-implementations rl-algorithms reinforcement-learning-environments mcts-algorithm mcts-agents
-
Updated
May 7, 2023 - Python
A Hex board game with a customizable Monte Carlo Tree Search (MCTS) agent with optional leaf parallelization in C++14. Includes a logging functionality for MCTS insights.
cpp14 mcts console-game monte-carlo-tree-search hex-game mcts-visualisation mcts-implementations mcts-algorithm console-game-engine mcts-agents mcts-solver leaf-parallelization mcts-hex
-
Updated
Jun 5, 2023 - C++
Tic-tac-toe/"noughts & crosses" written in Clojure (CLI + deps). AI powered by Monte Carlo tree search algorithm
board-game clojure reinforcement-learning functional-programming tic-tac-toe artificial-intelligence bitboards mcts reinforcement-learning-algorithms tictactoe noughts-and-crosses monte-carlo-tree-search artificial-intelligence-algorithms reinforcement-learning-agent mcts-implementations rl-algorithms reinforcement-learning-environments mcts-algorithm clojure-cli mcts-agents
-
Updated
Aug 19, 2022 - Clojure
Improve this page
Add a description, image, and links to the mcts-agents topic page so that developers can more easily learn about it.
Add this topic to your repo
To associate your repository with the mcts-agents topic, visit your repo's landing page and select "manage topics."