🔴 High Significance

Model Releases

🔴 HY-World 2.0: A Multi-Modal World Model for Reconstructing, Generating, and Simulating 3D Worlds — score 95 Sources: huggingface

We introduce HY-World 2.0, a multi-modal world model framework that advances our prior project HY-World 1.0. HY-World 2.0 accommodates diverse input modalities, including text prompts, single-view images, multi-view images, and videos, and produces 3D world representations. With text or single-view

🔴 How to Fine-Tune a Reasoning Model? A Teacher-Student Cooperation Framework to Synthesize Student-Consistent SFT Data — score 75 Sources: huggingface

A widely adopted strategy for model enhancement is to use synthetic data generated by a stronger model for supervised fine-tuning (SFT). However, for emerging reasoning models like Qwen3-8B, this approach often fails to improve reasoning capabilities and can even lead to a substantial drop in perfor

Developer Tools

🔴 DR^{3}-Eval: Towards Realistic and Reproducible Deep Research Evaluation — score 85 Sources: huggingface

Deep Research Agents (DRAs) aim to solve complex, long-horizon research tasks involving planning, retrieval, multimodal understanding, and report generation, yet their evaluation remains challenging due to dynamic web environments and ambiguous task definitions. We propose DR^{3}-Eval, a realistic a

🟡 Notable

Model Releases

🟡 Product Apr 17, 2026 Introducing Claude Design by Anthropic Labs Today, we’re launching Claude Design, a new Anthropic Labs product that lets you collaborate with Claude to create polished visual work like designs, prototypes, slides, one-pagers, and more. — score 50 Sources: lab_blog/Anthropic

Announcements Apr 7, 2026 Project Glasswing A new initiative that brings together Amazon Web Services, Anthropic, Apple, Broadcom, Cisco, CrowdStrike, Google, JPMorganChase, the Linux Foundation, Microsoft, NVIDIA, and Palo Alto Networks in an effort to secure the world's most critical software. Mar

🟡 Announcements Feb 4, 2026 Claude is a space to think We’ve made a choice: Claude will remain ad-free. We explain why advertising incentives are incompatible with a genuinely helpful AI assistant, and how we plan to expand access without compromising user trust. — score 50 Sources: lab_blog/Anthropic

Product Apr 17, 2026 Introducing Claude Design by Anthropic Labs Today, we’re launching Claude Design, a new Anthropic Labs product that lets you collaborate with Claude to create polished visual work like designs, prototypes, slides, one-pagers, and more. Announcements Apr 7, 2026 Project Glasswing

🟡 Dive into Claude Code: The Design Space of Today's and Future AI Agent Systems — score 45 Sources: huggingface

Claude Code is an agentic coding tool that can run shell commands, edit files, and call external services on behalf of the user. This study describes its comprehensive architecture by analyzing the publicly available TypeScript source code and further comparing it with OpenClaw, an independent open-

Developer Tools

🟡 RAD-2: Scaling Reinforcement Learning in a Generator-Discriminator Framework — score 65 Sources: huggingface

High-level autonomous driving requires motion planners capable of modeling multimodal future uncertainties while remaining robust in closed-loop interactions. Although diffusion-based planners are effective at modeling complex trajectory distributions, they often suffer from stochastic instabilities

🟡 GlobalSplat: Efficient Feed-Forward 3D Gaussian Splatting via Global Scene Tokens — score 55 Sources: huggingface

The efficient spatial allocation of primitives serves as the foundation of 3D Gaussian Splatting, as it directly dictates the synergy between representation compactness, reconstruction speed, and rendering fidelity. Previous solutions, whether based on iterative optimization or feed-forward inferenc

🟢 Incremental

Developer Tools

🟢 HiVLA: A Visual-Grounded-Centric Hierarchical Embodied Manipulation System — score 35 Sources: huggingface

While end-to-end Vision-Language-Action (VLA) models offer a promising paradigm for robotic manipulation, fine-tuning them on narrow control data often compromises the profound reasoning capabilities inherited from their base Vision-Language Models (VLMs). To resolve this fundamental trade-off, we p

🟢 ASGuard: Activation-Scaling Guard to Mitigate Targeted Jailbreaking Attack — score 25 Sources: huggingface

Large language models (LLMs), despite being safety-aligned, exhibit brittle refusal behaviors that can be circumvented by simple linguistic changes. As tense jailbreaking demonstrates that models refusing harmful requests often comply when rephrased in past tense, a critical generalization gap is re

🟢 UniDoc-RL: Coarse-to-Fine Visual RAG with Hierarchical Actions and Dense Rewards — score 15 Sources: huggingface

Retrieval-Augmented Generation (RAG) extends Large Vision-Language Models (LVLMs) with external visual knowledge. However, existing visual RAG systems typically rely on generic retrieval signals that overlook the fine-grained visual semantics essential for complex reasoning. To address this limitati

Infrastructure & Compute

🟢 LeapAlign: Post-Training Flow Matching Models at Any Generation Step by Building Two-Step Trajectories — score 5 Sources: huggingface

This paper focuses on the alignment of flow matching models with human preferences. A promising way is fine-tuning by directly backpropagating reward gradients through the differentiable generation process of flow matching. However, backpropagating through long trajectories results in prohibitive me

📄 New Papers

TitleCategoryScoreLink
HY-World 2.0: A Multi-Modal World Model for Reconstructing, Generating, and Simulating 3D Worldsmodel_release122Open
DR^{3}-Eval: Towards Realistic and Reproducible Deep Research Evaluationdeveloper_tool37Open
How to Fine-Tune a Reasoning Model? A Teacher-Student Cooperation Framework to Synthesize Student-Consistent SFT Datamodel_release36Open
RAD-2: Scaling Reinforcement Learning in a Generator-Discriminator Frameworkdeveloper_tool33Open
GlobalSplat: Efficient Feed-Forward 3D Gaussian Splatting via Global Scene Tokensdeveloper_tool27Open
BioHiCL: Hierarchical Multi-Label Contrastive Learning for Biomedical Retrieval with MeSH Labelscs.AI0Open
DALM: A Domain-Algebraic Language Model via Three-Phase Structured Generationcs.AI0Open
DataCenterGym: A Physics-Grounded Simulator for Multi-Objective Data Center Schedulingcs.AI0Open
PoInit-of-View: Poisoning Initialization of Views Transfers Across Multiple 3D Reconstruction Systemscs.AI0Open
Imperfectly Cooperative Human-AI Interactions: Comparing the Impacts of Human and AI Attributes in Simulated and User Studiescs.AI0Open
CLIMB: Controllable Longitudinal Brain Image Generation using Mamba-based Latent Diffusion Model and Gaussian-aligned Autoencodercs.AI0Open
VoodooNet: Achieving Analytic Ground States via High-Dimensional Random Projectionscs.AI0Open
Rethinking the Necessity of Adaptive Retrieval-Augmented Generation through the Lens of Adaptive Listwise Rankingcs.AI0Open
Conjunctive Prompt Attacks in Multi-Agent LLM Systemscs.AI0Open
HYPERHEURIST: A Simulated Annealing-Based Control Framework for LLM-Driven Code Generation in Optimized Hardware Designcs.AI0Open

🏢 Lab Blog Posts