今日从 arXiv 订阅中筛选 8 篇论文。

Arxiv ID2605.10564幻觉翻译2605.10564

⚡ DeepSight: Driving World Model via BEV Latent Prediction

DeepSight: Driving World Model via BEV Latent Prediction
Arxiv ID2605.10426幻觉翻译2605.10426

⚡ CoWorld-VLA: Multi-Expert World Model for Autonomous Driving

Arxiv ID2605.08830幻觉翻译2605.08830

⚡ VECTOR-Drive: VLA Expert Routing with Qwen2.5-VL-3B

Arxiv ID2605.09701幻觉翻译2605.09701

⚡ DriveFuture: Future-Aware Latent World Model

DriveFuture: Future-Aware Latent World Model
Arxiv ID2605.08735幻觉翻译2605.08735

⚡ CollabVR: VLM-VGM Collaborative Video Reasoning

Arxiv ID2605.10858幻觉翻译2605.10858

⚡ WorldLens: Driving World Model Benchmark

WorldLens: Driving World Model Benchmark
Arxiv ID2605.09422幻觉翻译2605.09422

⚡ Perception Without Engagement: Causal Discovery Deficit in LMMs

Perception Without Engagement: Causal Discovery Deficit in LMMs
Arxiv ID2605.09650幻觉翻译2605.09650

⚡ Workspace Optimization for Agent Training

Workspace Optimization for Agent Training

自动生成于 2026-05-13 · 基于 arXiv Daily Digest