The papers

Most central to this idea first, not most recent.

• Big Techcs.CL

  Agentic document QA pays off mainly with stronger vision-language models

  Amazon.com · Sept 2026

• Top Universitycs.RO

  Visual action rehearsal helps language models control robot manipulation

  HKUST(GZ), CUHK · Sept 2026

• Big Techcs.CV

  A single visual agent spans mobile, desktop, web, and tool use

  Apple · Sept 2026

• Industrycs.AI

  Decouple high-level VLM planning from low-level execution to speed up mobile agents

  Rice University · Sept 2026

• Independentcs.RO

  A semantic harness lets vision-language models control different robots

  Sept 2026

• Independentcs.AI

  Scene-grounded decoding keeps vision-language plans executable and visually supported

  Sept 2026

• Industrycs.AIcode

  Agents learn selectively from imperfect vision-language model teachers

  Fondazione Bruno Kessler, University of Torino · Sept 2026

• Top Universitycs.CV

  Factorized reinforcement learning improves vision-language models’ spatial reasoning

  Joy Future Academy, The Hong Kong University of Science and Technology (Guangzhou) · Sept 2026

• Independentcs.CV

  Coding agents combine generated imagery with editable web-based visual layouts

  Sept 2026

• Chinese Techcs.CV

  Conversational image editing agent learns interpretable tool use

  Shanghai Innovation Institution, Huawei Technologies Ltd. · June 2026

• Independentcs.SE

  Multi-image evidence can help models repair software, but unreliably

  Sept 2026

• Chinese Techcs.AIcode

  C3M preserves multimodal evidence across sessions within fixed memory budgets

  Huzhou Normal University, Alibaba Group · Sept 2026

• Top Universitycs.RO

  Saliency-driven workspace tokens give robots lightweight task memory

  Massachusetts Institute of Technology, Carnegie Mellon University · Sept 2026

• Chinese Techcs.AI

  Self-improving context programs help frozen models understand long videos

  Tencent · Aug 2026

• Top Universitycs.CVcode

  MLLMs fail to sustain goal-directed navigation in real-scale 3D city

  Shanghai Jiao Tong University, National University of Singapore · Aug 2026

• Industrycs.CL

  Self-evolving loop synthesizes high-quality multimodal training data

  vivo AI Lab · Aug 2026

• Chinese Techcs.CV

  Adaptive tool use improves video research agents’ accuracy and efficiency

  Accio Team, Alibaba Group, Beijing Key Laboratory of Intelligent Information Technology · Aug 2026

• Big Techcs.AI

  New framework lets you train AI agents inside the same harness systems they use at inference

  Columbia University, Dartmouth College · July 2026

• Big Techcs.AI

  Skill evolution improves agent performance in image generation workflows

  University of Pennsylvania, Nvidia · July 2026

• Research Labcs.CR

  Execution Logs Can Mislead Visual Judges in Video-Generation Agents

  RIKEN · Sept 2026

• Independentcs.CVcode

  Image-only pretraining helps build a strong open image generator

  Sept 2026

• Chinese Techcs.CV

  Streaming video memory works better when internalized as evolving latent tokens

  Nanjing University of Science and Technology, Ant Group · Sept 2026

• Top Universitycs.CVcode

  Multi-agent framework coordinates narrative and visual consistency for long-form video

  The University of Hong Kong, South China University of Technology · June 2026

• Independentcs.CV

  Code-driven agents generate images via sketching and texturing stages

  May 2026

• Chinese Techcs.CV

  Unified visual-generation agentic model outperforms larger closed-source models

  Tencent Hunyuan, Hong Kong University of Science and Technology · Mar 2026