ICML 2026 index

A classified index map of ICML 2026 — talks and papers with links to code and keywords. Deep write-ups are added over time and linked from here.

The index map for ICML 2026 (Seoul · COEX · July 6–11). Each row links out to arXiv / code and carries keywords + a topic. Deep reading notes are written for selected entries and linked in the Post column; the running day-by-day archive lives in the ICML 2026 — live archive post.

Invited talks

Speaker Affiliation Talk Topic
       
Pascale Fung HKUST; Director, CAiRE TBA agentic-ai
       
Susan Athey Stanford GSB Causal Inference with Transformer Models
Tue Jul 7, 16:00 KST, Hall C
causal-inference
       
Sham M. Kakade Harvard; Kempner Institute TBA ml-theory
       
Aviv Regev Genentech (EVP, Research & Early Development) TBA computational-biology
       
Verena Rieser Google DeepMind TBA nlp-dialogue
       
Arvind Narayanan Princeton; Director, CITP TBA ai-policy
       

Papers

robotics-vla

Paper Authors Links Keywords Post
         
Qwen-VLA: Unifying Vision-Language-Action Modeling across Tasks, Environments, and Robot Embodiments Qwen Team arXiv · code vla · dit-decoder · flow-matching · embodiment-aware · cross-embodiment note
         
Spatial Memory for Out-of-Vision Manipulation in Vision-Language-Action Pengteng Li, Weiyu Guo, He Zhang, et al.   vla · spatial-memory · out-of-vision · manipulation —
         
         

How this grows. New entries are appended to _data/icml2026.yml (the index), classified by topic. The full accepted-paper list is pulled in batches from PMLR / OpenReview. A selected entry graduates into a full reading note (the post: link) — the index stays the map; the posts are the territory.