AI Research Brief
Search
Methodology
中文
Swap the Action Interface, Gain 11 Points on Spatial Reasoning
23 selected from 343 papers
Featured
SpatialClaw: Rethinking Action Interface for Agentic Spatial Reasoning
score 10
入选 HF Daily Papers; HF 热度: 80 upvotes (+4); 有代码实现; 关键词(3): agentic, reasoning, vision-language
MoVerse: Real-Time Video World Modeling with Panoramic Gaussian Scaffold
score 9
入选 HF Daily Papers; HF 热度: 10 upvotes (+3); 有代码实现; 关键词(3): latency, real-time, reasoning
InterleaveThinker: Reinforcing Agentic Interleaved Generation
score 10
入选 HF Daily Papers; HF 热度: 74 upvotes (+4); 有代码实现; 关键词(5): GRPO, agentic, reasoning, embodied, open-source
HarnessBridge: Learnable Bidirectional Controller for LLM Agent Harness
score 8
入选 HF Daily Papers; HF 热度: 9 upvotes (+2); 有代码实现; 关键词(2): lightweight, instruction tuning
RepWAM: World Action Modeling with Representation Visual-Action Tokenizers
score 7
入选 HF Daily Papers; HF 热度: 5 upvotes (+2); 有代码实现
EurekAgent: Agent Environment Engineering is All You Need For Autonomous Scientific Discovery
score 10
入选 HF Daily Papers; HF 热度: 21 upvotes (+4); 有代码实现; 关键词(1): open-source
LabVLA: Grounding Vision-Language-Action Models in Scientific Laboratories
score 10
入选 HF Daily Papers; HF 热度: 50 upvotes (+4); 有代码实现; 关键词(2): pretraining, vision-language
VideoMDM: Towards 3D Human Motion Generation From 2D Supervision
score 8
入选 HF Daily Papers; HF 热度: 16 upvotes (+3); 有代码实现
Also Worth Noting
Demystifying Hidden-State Recurrence: Switchable Latent Reasoning with On-Policy Reinforcement Learning
score 7
入选 HF Daily Papers; HF 热度: 18 upvotes (+3); 关键词(2): GRPO, reasoning
Zero-source LLM Hallucination Detection with Human-like Criteria Probing
score 4
关键词(1): reasoning; 顶会接收: ICML
Self-Guidance: Enhancing Neural Codecs via Decoder Manifold Alignment
score 4
关键词(3): lightweight, quantization, coding; 顶会接收: ICML
TWLA: Achieving Ternary Weights and Low-Bit Activations for LLMs via Post-Training Quantization
score 4
关键词(4): compression, quantization, deployment, post-training; 顶会接收: ICML
Functional Cache Grafting: Robust and Rapid Code-Policy Synthesis for Embodied Agents
score 4
关键词(2): latency, embodied; 顶会接收: ICML
Reasoning for Mobile User Experience with Multimodal LLMs: Task, Benchmark, and Approach
score 4
关键词(3): latency, code generation, reasoning; 顶会接收: CVPR
Cross-Modal Masked Compositional Concept Modeling for Enhancing Visio-Linguistic Compositionality
score 4
关键词(2): vision-language, text-to-image; 顶会接收: ACL
PolyFlow: Safe and Efficient Polytope-Constrained Flow Matching with Constraint Embedding and Projection-free Update
score 4
关键词(1): latency; 顶会接收: ICML
MaxProof: Scaling Mathematical Proof with Generative-Verifier RL and Population-Level Test-Time Scaling
score 8
入选 HF Daily Papers; HF 热度: 72 upvotes (+4); 关键词(1): scaling
Budget-Constrained Step-Level Diffusion Caching
score 4
关键词(2): deployment, latency; 顶会接收: ICML
Towards Effective Waste Segmentation for Automated Waste Recycling in Cluttered Background
score 4
关键词(1): production; 顶会接收: ICML
Multi-Agent Reinforcement Learning from Delayed Marketplace Feedback for Objective-Weight Adaptation in Three-Sided Dispatch
score 4
关键词(1): production; 顶会接收: ICML
Surflo: Consistent 3D Surface Flow Model with Global State
score 4
入选 HF Daily Papers; HF 热度: 2 upvotes (+1)
CLARITree: Cholesky and Lookahead Accelerations for Regression with Interpretable Piecewise Linear Trees
score 3
顶会接收: ICML
Layer-Resolved Optimal Transport for Hallucination Detection in NMT and Abstractive Summarization
score 3
顶会接收: ICML