Paper Tracking

An automated weekly digest. Every Monday an agent scans arxiv and HuggingFace Daily Papers, downloads the PDFs, and summarizes the top 3 picks per topic.

Week of June 29, 2026 · View archive →

Graph × LLM

Graph-structured reasoning, graph foundation models, and LLM-augmented GNNs.

Physics Question Scene Graph: Fine-grained Evaluation of Physical Plausibility in Text-to-Video Generation

↑ 1 📚 20084 ★ 0 Jun 23

Problem: Video generation models are increasingly capable of producing realistic videos, but they still struggle to generate videos that follow basic physical laws. Comp
Model:
Code: atinpothiraj/pqsg

A Fair Evaluation of Graph Foundation Models for Node Property Prediction

↑ 0 📚 1890 ★ 0 Jun 23

Problem: Due to the wide use of graph-structured data in different fields of industry and science, the development of Graph Foundation Models (GFMs) has recently attract
Model:
Code: not released

Understanding Rollout Error in Graph World Models

↑ 0 ★ 0 Jun 26

Problem: World models are often used for planning by rolling learned dynamics forward. Many planning environments, however, are not vectors or images; they are graphs of
Model:
Code: Hik289/graph_

Vision-Language Models

Multimodal VLMs, vision encoders, and image-text foundation models.

Wan-Streamer v0.1: End-to-end Real-time Interactive Foundation Models

↑ 98 📚 83 ★ 0 Jun 22

Problem: We present Wan-Streamer, a native-streaming, end-to-end interactive foundation model designed from the ground up for real-time, low-latency, full-duplex audio-v
Model:
Code: not released

ShutterMuse: Capture-Time Photography Guidance with MLLMs

↑ 45 📚 7140 ★ 0 Jun 23

Problem: Real-world photography requires capture-time guidance for both camera framing and subject pose. Yet existing aesthetic cropping benchmarks mainly evaluate post-
Model:
Code: not released

ViQ: Text-Aligned Visual Quantized Representations at Any Resolution

↑ 38 📚 7100 ★ 0 Jun 24

Problem: A unified representation for text and vision is a natural pursuit, as it enables simpler multimodal modeling and more efficient training. However, representing
Model:
Code: yuxumin/ViQ

World Models

Predictive world models, JEPA-style learning, video generation, and embodied simulators.

Qwen-AgentWorld: Language World Models for General Agents

↑ 139 📚 554 ★ 0 Jun 23

Problem: A world model predicts environment dynamics based on current observations and actions, serving as a core cognitive mechanism for reasoning and planning. In this
Model:
Code: expressjs/express.git

DomainShuttle: Freeform Open Domain Subject-driven Text-to-video Generation

↑ 64 📚 8106 ★ 0 Jun 23

Problem: Open domain subject-driven text-to-video (S2V) generation has drawn significant interest in academia and industry. Open domain S2V mainly involves two scenarios
Model:
Code: not released
⚠ Interested, but agent could not fetch the PDF — summary based on abstract only.

Hallucination in World Models is Predictable and Preventable

↑ 8 📚 21396 ★ 0 Jun 25

Problem: Modern generative world models render increasingly realistic action-controllable futures, yet they frequently hallucinate: rollouts remain visually fluent while
Model:
Code: google-research/robodesk

Spatial Single-Cell Study

Single-cell + spatial omics foundation models, virtual cells, biological world models, and biomedical digital twins.

SP-Mind: An Autonomous Reasoning Agent for Spatial Proteomics Analysis

↑ 0 📚 95 ★ 0 Jun 23

Problem: Spatial proteomics enables single-cell-resolution characterization of protein expression within tissue architecture, playing a critical role in understanding tu
Model:
Code: not released