Skip to content

NeurIPS 2026 — Program Notes (pre-meeting prep)

Internal working file. Today is May 11, 2026 — the meeting is ~7 months out. Main paper notifications go Sep 24; workshop accept/reject Sep 29; accepted-paper lists post on OpenReview Oct–Nov. No accepted-paper data exists yet. This file holds the scaffolding so we can move fast once the lists drop.

Verified facts

Field Value Source
Conference 40th Conference on Neural Information Processing Systems neurips.cc
Model three-city satellite (first time at three hubs) neurips.cc
Sydney dates Dec 6–12, 2026 neurips.cc/Conferences/2026
Atlanta dates Dec 9–13, 2026 neurips.cc/Conferences/2026
Paris dates Dec 9–13, 2026 neurips.cc/Conferences/2026
Paper submission deadline May 6, 2026 (AoE) — past neurips.cc
Workshop application deadline Jun 6, 2026 (AoE) CallForWorkshops
Workshop accept notifications Jul 11, 2026 (AoE) CallForWorkshops
Paper author notifications Sep 24, 2026 neurips.cc
Workshop final-decision deadline Sep 29, 2026 CallForWorkshops
Workshop suggested-contribution deadline Aug 29, 2026 CallForWorkshops
License CC-BY (OpenReview default) OpenReview

Sources we will rely on

Source When usable Coverage
NeurIPS 2026 home now dates, satellite hubs, organizers
Call for Workshops now workshop timeline
OpenReview NeurIPS group from Sep 24 accepted main-track papers
Per-workshop OpenReview venues from Oct–Nov workshop accepted papers
SlidesLive (neurips) from ~Dec 14 recorded orals + workshop talks
bioRxiv / arXiv continuous preprints, filterable by "to appear at NeurIPS 2026"
#NeurIPS2026 on Bluesky / Mastodon / X Dec 6–13 real-time talk reactions
Lab blogs (DeepMind, Anthropic, Insitro, Recursion, Genentech prescient design, Isomorphic Labs) rolling preprints + summary posts of NeurIPS papers from these labs

Filter criteria (locked — mirrors tools/index.md)

A paper enters this vault only if it matches one of three filters. Reproduced here for the pre-meeting screening pass.

Filter 1 — workshop track

Accept anything in these workshops (canonical short names; 2026 roster TBD Sep 29):

  • MLSB — Machine Learning in Structural Biology
  • AI4NewDrugs / AI4D3 — AI for New Drug Modalities / AI for Drug Discovery
  • LMRL — Learning Meaningful Representations of Life
  • GenAI4Health — Generative AI for Health
  • AIM-FM — Advancements In Medical Foundation Models
  • AI4Science — AI for Science (bio / chem subset only)

Filter 2 — main-track keyword

Run this regex over titles + abstracts once the main-track list drops:

(cancer|oncolog|tumor|metasta|carcinom|melanoma|leukemi|lymphoma)
| (genomic|epigenom|transcriptom|proteom|metabolom|methyl)
| (single[- ]cell|scRNA|scATAC|spatial transcriptom|spatial omic)
| (perturbation|CRISPR|knockout|knockdown|sgRNA)
| (drug discover|target ident|molecular property|ADMET|small molecul|biologic|de novo design|generative chemistry)
| (protein structur|protein function|protein language|equivariant|RFdiffusion|AlphaFold|Boltz|ESM)
| (foundation model.*(biolog|cell|genom|tissue|pathol|clinical|medical))
| (electronic health record|EHR|clinical NLP|medical imaging|pathology|histolog|radiolog|diagnos)
| (virtual cell|cell atlas|cellular state)

Filter 3 — agentic-AI / LLM-for-science subset

Independent of bio framing, accept any main-track or workshop paper on:

  • LLM agents with tool-use, code-execution, or experiment-design loops
  • Scientific-reasoning benchmarks (SWE-bench-class, sci-agent-class, lab-task benchmarks)
  • Autonomous research / hypothesis-generation systems
  • Multi-agent scientific workflows

Rationale: these adapt to oncology within 6 months (the empirical pattern from 2024 → AACR 2025 and 2025 → AACR 2026) and feed the AACR agentic-AI axis directly.

Expected paper themes (5 anchors)

Pre-meeting predictions, to be reconciled against the actual list in September. These are the buckets we expect tools/ entries to cluster into.

Theme 1 — Foundation models for biology, second generation

scGPT, Geneformer, and ESM are 2022–2024 vintage. NeurIPS 2026 should feature the next-gen models: cross-modal (DNA + RNA + protein + tissue + clinical), multi-million-cell training corpora, instruction-tuned for downstream tasks (perturbation prediction, virtual-screen rerank), and benchmarked on virtual-cell tasks rather than embedding-quality proxies. Workshop home: LMRL, with main-track spillover.

Theme 2 — Agentic AI for science

Tool-use LLMs that drive bioinformatics pipelines, design wet-lab experiments, or autonomously navigate clinical-decision trees. 2025 surfaced ChemCrow / Coscientist-class systems; 2026 should bring oncology-specific agents (tumor-board agents, trial-matching agents, multi-omics interpretation agents). Workshop home: main-track + GenAI4Health + AIM-FM.

Theme 3 — Single-cell + spatial ML methods

Cell-state representation learning, perturbation-response prediction (replogle-class, sci-Plex-class benchmarks), spatial-transcriptomics imputation and cell-cell communication inference, virtual-tissue generation. Workshop home: LMRL; main-track for the methods-heavy contributions.

Theme 4 — Generative chemistry and protein design for oncology targets

De novo small-molecule generation conditioned on oncology targets, antibody / TCR / peptide design, equivariant diffusion models for protein–ligand docking, RFdiffusion descendants, Boltz / AlphaFold3-class structure models. Workshop home: MLSB + AI4D3.

Theme 5 — Medical foundation models and clinical NLP

Pathology FMs (CHIEF / GigaPath / Virchow successors), radiology FMs, EHR-derived patient-trajectory models, clinical-trial-matching language models, regulatory and safety work on medical LLMs. Workshop home: AIM-FM + GenAI4Health; main-track for the biggest model releases.

Pre-meeting timeline (action items)

Date Action
Now → Sep 24, 2026 passive monitoring; no extraction
Sep 24, 2026 main-track accepted list drops on OpenReview — run keyword regex, build candidate list
Sep 29, 2026 workshop roster finalized — confirm which of the six bio/health workshops actually run, identify any new entrants
Oct–Nov 2026 per-workshop accepted-paper lists drop — extend candidate list
Nov 2026 scaffold likely tools/ pages from titles + bioRxiv mirrors (no waiting for the talks)
Dec 6–13, 2026 meeting week — opportunistic real-time capture from social, especially agentic-AI orals
Mid-Dec 2026 → Jan 2027 bulk tools/ population from SlidesLive recordings + camera-ready papers

Caveats

  • No accepted-paper list exists yet. Everything above is anchored on past-year patterns. The 2026 workshop roster could drop a recurring bio workshop or add a new one we haven't planned for; the keyword filter is the safety net.
  • Three-hub model is new. It's plausible that workshop programs differ across Sydney / Atlanta / Paris. Until the per-hub programs publish (~Oct), we assume a unified program; revisit if not.
  • OpenReview lag. Camera-ready and supplementary materials sometimes lag the accept notification by weeks. Tool pages may need a "weights / code TBA" placeholder for the first month.
  • Agentic-AI scope creep risk. Filter 3 is broader than Filters 1–2 by design, but it can swell if applied loosely. Keep the "science / tool-use benchmark required" caveat from tools/index.md open question 2.