https://github.com/facebookresearch/decrypto
공식 코드The Decrypto Benchmark for Multi-Agent Reasoning and Theory of Mind
태스크의 하위 분류와 논문·코드·평가·기법·데이터셋을 탐색하세요.
관련 논문과 평가 결과에 연결된 구현입니다. 출처에서 확인된 저자 구현에는 공식 코드 표시가 있습니다.
The Decrypto Benchmark for Multi-Agent Reasoning and Theory of Mind
DD-PPO: Learning Near-Perfect PointGoal Navigators from 2.5 Billion Frames
VER: Scaling On-Policy RL Leads to the Emergence of Navigation in Embodied Rearrangement
Memory-Augmented Reinforcement Learning for Image-Goal Navigation
Talk the Walk: Navigating New York City through Grounded Dialogue
Quantifying Adaptability in Pre-trained Language Models with 500 Tasks
TrajSceneLLM: A Multimodal Perspective on Semantic GPS Trajectory Analysis
Instance-level quantitative saliency in multiple sclerosis lesion segmentation
Fast-Slow Test-Time Adaptation for Online Vision-and-Language Navigation
BabelBench: An Omni Benchmark for Code-Driven Analysis of Multimodal and Multistructured Data
TextGames: Learning to Self-Play Text-Based Puzzle Games via Language Model Reasoning
Grid Cell-Inspired Fragmentation and Recall for Efficient Map Building
Discovering Sensorimotor Agency in Cellular Automata using Diversity Search
Decomposed Soft Prompt Guided Fusion Enhancing for Compositional Zero-Shot Learning
Waypoint Generation in Row-based Crops with Deep Learning and Contrastive Clustering
Generating by Understanding: Neural Visual Generation with Logical Symbol Groundings
DeepResearcher: Scaling Deep Research via Reinforcement Learning in Real-world Environments
Using stigmergy as a computational memory in the design of recurrent neural networks
Infogent: An Agent-Based Framework for Web Information Aggregation
MagicEye: An Intelligent Wearable Towards Independent Living of Visually Impaired
Learning future terrorist targets through temporal meta-graphs
M2A: Motion Aware Attention for Accurate Video Action Recognition
Text2Robot: Evolutionary Robot Design from Text Descriptions
Empower Nested Boolean Logic via Self-Supervised Curriculum Learning
출처: Papers with Code · CC BY-SA 4.0. 아카이브 자료와 새로 수집한 논문은 별도로 관리합니다.