今日从 arXiv 订阅中筛选 10 篇论文。

Arxiv ID2607.29586幻觉翻译2607.29586

⚡ TraceViT: Grounded Trace Supervision for Visual Abstract Reasoning

**Relevance:** 4.4 | **机构:** Zhejiang University / Zhejiang Lab

TraceViT: Grounded Trace Supervision for Visual Abstract Reasoning
Arxiv ID2607.28638幻觉翻译2607.28638

⚡ Learning Stateful Predictive Knowledge From Experience

**Relevance:** 3.9

Arxiv ID2607.29237幻觉翻译2607.29237

⚡ CorrelationFlow: A Training-Free Geometric Approach for LiDAR Scene Flow Estimation

**Relevance:** 3.7 | **机构:** Nantes / Nottingham / Queensland / TU Delft

CorrelationFlow: A Training-Free Geometric Approach for LiDAR Scene Flow Estimation
Arxiv ID2607.29627幻觉翻译2607.29627

⚡ FlexComposer: Unified Video Compositing from Images to Dynamic Footage with Flexible Trajectory Control

**Relevance:** 4.0

FlexComposer: Unified Video Compositing from Images to Dynamic Footage with Flexible Trajectory Control
Arxiv ID2607.29401幻觉翻译2607.29401

⚡ OSEF: One-Step Evidence Fusion for Cross-Video Scene Procedure Planning

**Relevance:** 4.1

Arxiv ID2607.28678幻觉翻译2607.28678

⚡ ViSAGE: Constructing Self-Correcting Memories for Long-Form Video Understanding

**Relevance:** 3.7

Arxiv ID2607.28991幻觉翻译2607.28991

⚡ CAER: Conflict-Aware Evidence Routing with Dual Prefix Experts for Multimodal Large Language Models

**Relevance:** 3.8

CAER: Conflict-Aware Evidence Routing with Dual Prefix Experts for Multimodal Large Language Models
Arxiv ID2607.29180幻觉翻译2607.29180

⚡ MoRAE: Flow-Friendly Self-Supervised Latents for Text-to-Motion Generation

**Relevance:** 4.1

Arxiv ID2607.29185幻觉翻译2607.29185

⚡ Learning Latent Reasoning Traces for Scalar Reward Models End-to-End

**Relevance:** 3.9

Learning Latent Reasoning Traces for Scalar Reward Models End-to-End
Arxiv ID2607.28665幻觉翻译2607.28665

⚡ Sensitivity Analysis of GRU, LSTM and Transformer Encoder in Classification of Automated Driving Systems

**Relevance:** 3.4 | **机构:** University of Memphis

自动生成于 2026-08-03 · 基于 arXiv Daily Digest