Reading List

On My Radar

Papers

11
01

Learning to Discover at Test Time

TTT-Discover Team · arXiv

02

Qwen3-TTS Technical Report

Qwen Team, Alibaba Group · arXiv

03

LLM-in-Sandbox Elicits General Agentic Intelligence

Tsinghua University, Renmin University et al. · arXiv

04

Cosmos Policy: Fine-Tuning Video Models for Visuomotor Control and Planning

Stanford University, NVIDIA · arXiv

05

Scaling Text-to-Image Diffusion Transformers with Representation Autoencoders

New York University · arXiv

06

Controlling Long-Horizon Behavior in Language Model Agents with Explicit State Dynamics

Independent Researcher · arXiv

07

BayesianVLA: Bayesian Decomposition of Vision Language Action Models via Latent Action Queries

BayesianVLA Team · arXiv

08

EvoCUA: Evolving Computer Use Agents via Learning from Scalable Synthetic Experience

Meituan, Tongji University et al. · arXiv

09

You Need Better Attention Priors

Stanford University · arXiv

10

Toward Efficient Agents: Memory, Tool learning, and Planning

Shanghai AI Lab, Fudan University · arXiv

11

Agentic Reasoning for Large Language Models

Survey Authors · arXiv

Editoriali

27