πŸ”¬ Into medicine too? Top papers, preprints & trials, every morning β˜•Distill Medicine β†’
AI research, distilled.

500+ AI papers drop daily.
You read the ones that matter.

Your personalized feed of AI papers, code, and models β€” from arXiv, NeurIPS, CVPR, Nature, GitHub, HuggingFace πŸ€—, and more β€” tailored to your interests, plus follow the researchers you care about. It all lands in your inbox every morning, with a personalized daily AI news feed. Then chat with any paper to dig deeper, in seconds.

Free to start

Used by researchers and engineers at

University of OxfordUniversity of CambridgeHarvard Medical SchoolMassachusetts Institute of Technology
CaltechUniversity of EdinburghSamsungHuawei

Pick your topics β€” or name your own ✨

Choose from 86 specializations β€” and add any custom topic you want to track. Custom topics show first in your feed.

Popular specializations
Select All Topics (86)
Large Language ModelsComputer VisionDiffusion ModelsReinforcement LearningMultimodal LearningFoundation ModelsNLPMachine LearningRetrieval-Augmented GenerationObject Detection
Browse all 86 specializationsβ–Ό
✨ Track a custom topic
MRI reconstruction|
Add
✨ MRI reconstruction Γ—βœ¨ Protein structure prediction Γ—
πŸ‘€ Follow researchers
Yann LeCun|
Yann LeCun
New York University, Meta AI Β· 600+ papers
πŸ‘€ Yann LeCun Γ—

Your morning research feed β˜•

Ask MyResearch
Sample FeedNext feed in 16h 42m
Papers
Search
Bookmarks (3)
Collections
Community
πŸ‘€ From researchers you follow
πŸ‘€ YANN LECUNarXiv Β· 3d
A Path Towards Autonomous Machine Intelligence
Y. LeCun
πŸ‘€ YANN LECUNarXiv Β· 12d
Self-Supervised Learning from Images with a Joint-Embedding Predictive Architecture (I-JEPA)
M. Assran, Q. Duval, I. Misra … Y. LeCun
✨ Your custom topics
✨ MRI RECONSTRUCTIONIEEE TMI · 5h
Diffusion Priors for Accelerated MRI Reconstruction from Undersampled k-Space
S. Nakamura, R. Patel, et al.
Computer Vision4h agoCVPR 2026CODEπŸ”₯ HOT
RDFace: A Benchmark Dataset for Rare Disease Facial Image Analysis under Extreme Data Scarcity and Phenotype-Aware Synthetic Generation β†—
Feng, Long, Ali, Lou et al.
SUMMARY Curated benchmark of 456 pediatric facial images across 103 rare genetic conditions for evaluating AI diagnosis under extreme data scarcity. Phenotype-aware synthetic augmentation via DreamBooth and FastGAN improves diagnostic accuracy by up to 13.7%.
Their DreamBooth augmentation improved accuracy by 13.7% β€” check if this phenotype-filtering approach works for our rare condition dataset.
Chat with paper β†’Read β†’Save to Zotero+ List
Large Language Models6h agoNeurIPS 2025CODEGITHUBπŸ”₯ HOT
FlashPrefill: Instantaneous Pattern Discovery and Thresholding for Ultra-Fast Long-Context Prefilling β†—
Fan, Huang, Wu, Wang, He
SUMMARY Sparse attention framework that discovers vertical, slash, and block patterns instantaneously for ultra-fast LLM prefilling. Achieves 27.78x speedup on 256K sequences while maintaining 1.71x gains even at 4K context length.
Chat with paper β†’Read β†’Save to Zotero+ List
Computer Vision8h agoCVPR 2026CODE
InfiniDepth: Arbitrary-Resolution and Fine-Grained Depth Estimation with Neural Implicit Fields β†—
Liu, Zhang, FermΓΌller, Choi et al.
Chat with paper β†’Read β†’Save to Zotero+ List
Trending in AI
TRENDING5h agoCODEGITHUB
Vero: An Open RL Recipe for General Visual Reasoning β†—
Sarch, Cai, Wang, Wu et al. Β· Princeton University
SUMMARY Fully open VLM family trained with single-stage RL on a 600K-sample dataset spanning 59 datasets and six visual reasoning categories. Outperforms Qwen3-VL-8B-Thinking on 23 of 30 benchmarks without any proprietary training data.
Chat with paper β†’Read β†’Save to Zotero+ List
Trending on GitHub
GITHUBPython2d ago⭐ 8.2K
huggingface/lerobot
State-of-the-art AI for real-world robotics in PyTorch
Trending on HuggingFace πŸ€—
πŸ€— MODELimage-text-to-text1d ago❀ 4.8K
microsoft/Florence-2-large
Unified vision foundation model for captioning, detection, and grounding
πŸ“° Your Daily AI News
OpenAI releases new frontier reasoning model
TechCrunch Β· 2h agoLLMs
New diffusion model tops image-generation benchmark
The Verge Β· 5h agodiffusion
⏰ Deadlines
~80 upcoming
MLHC 2026 β†—
Apr 30, 2026
1d
NeurIPS 2026 β†—
May 22, 2026
22d
SIGGRAPH Asia 2026 β†—
May 28, 2026
28d
CoRL 2026 β†—
Jun 30, 2026
61d
BMVC 2026 β†—
Jul 4, 2026
65d
ML4H 2026 β†—
Jul 31, 2026
92d
BIBM 2026 β†—
Aug 22, 2026
114d
WSDM 2027 β†—
Aug 8, 2026
100d
Scroll for the full list. Always verify on the conference site.
πŸŽ“ Events & Schools
40+ schools
SummerMARSS β†—
Mar 23, 2026
3d
SummerMedICSS β†—
Apr 24, 2026
35d
SummerMLSS β†—
Summer (varies by year)
site
SummerEEML β†—
Summer
site
SummerICVSS β†—
July
site
WinterVLPR β†—
Winter
site
Summer/winter schools, symposia & workshops.

Search by meaning, not keywords πŸ”

Describe what you want in plain language β€” Distill finds matching papers across the entire research literature by meaning, then lets you chat, summarize, bookmark, or save them to a list.

transformers for brain tumor segmentation in MRI
Search
96% matcharXiv Β· 2021
TransBTS: Multimodal Brain Tumor Segmentation Using Transformer
W. Wang, C. Chen, M. Ding, J. Li et al.
Chat with paper β†’SummaryRead β†’Share+ List
91% matcharXiv Β· 2023
3D Brainformer: 3D Fusion Transformer for Brain Tumor Segmentation
R. Nian, G. Zhang, Y. Sui, S. K. Warfield et al.
Chat with paper β†’SummaryRead β†’Share+ List

Chat with any paper πŸ’¬

Read the paper, ask the paper β€” side by side. Grounded in the paper's full text, with citations to specific sections.

Chat with paper
Compositional Generalization in Vision-Language Models
2/50 questions
What is the key contribution?
Key Contribution
The paper introduces a new benchmark for evaluating how vision-language models handle attribute-object combinations they did not see during training.
Two main findings:
  • State-of-the-art models drop 30–50% on novel combinations.
  • A contrastive training objective closes ~60% of that gap.
Ask another question...
Compositional Generalization in Vision-Language Models
A. AuthorsΒΉ, B. AuthorsΒ², C. AuthorsΒΉ Β· ΒΉUniversity Lab Β· Β²Research Org
Abstract
1. Introduction
2. Methodology
Page 1 of 14

MyResearch πŸ“š

Not one paper β€” all your bookmarked papers. Ask across everything you've saved: compare papers, find connections, surface what's relevant, and draft related-work with citations.

MyResearch
3/15 questions this month
What do the papers I saved this month agree on about diffusion models?
Across your 6 saved papers
Three converge on a clear point:
  • Classifier-free guidance lifts sample quality but reduces diversity β€” in Scalable Diffusion and Guided Generation.
  • Latent-space diffusion cuts compute ~10Γ— with little quality loss (Latent Diffusion).
Where they differ: Scalable Diffusion favors a cosine noise schedule, while Elucidating Design argues for a learned one.
Ask about your bookmarks...

Sources

Every top venue. One feed.

Papers, code, and models from the best conferences, journals, and repositories in AI.

Conferences
NeurIPSICMLICLRCVPRECCVICCVACLEMNLPAAAI
Journals
NatureNature MIScienceTPAMIJMLR
Preprints
arXiv
Code & Repositories
GitHubHuggingFace πŸ€—+ more

Features

Everything distilled. Nothing wasted.

Personalized Feed

Papers matched to your interests across 86 specializations. Updated every morning.

New Since Last Visit

Come back to a feed that knows what you’ve already seen β€” a count and markers highlight every paper published since your last visit, so you never re-scan the same list.

PRO

Custom Topics

Track any niche you name β€” like 'MRI reconstruction' β€” not just fixed categories. We expand it with synonyms across every source we cover, then surface it first in your feed.

PRO

Follow Researchers

Follow the people whose work you can’t miss β€” get an in-app + email alert the moment they publish, and their newest papers surface at the top of your feed.

Daily AI News

Daily AI news from across the web β€” funding rounds, product launches, model releases, policy. Free for everyone; Pro tunes it to your topics with a headline count you choose.

Share Your Paper

Upload your own paper β€” it’s shared with researchers on Distill AI who follow your topics, and you get a beautiful page with an auto-generated social card to post anywhere.

PRO

My Library

Upload your own PDFs β€” private to you. Chat with them, get summaries, organize into reading lists, and chat across a whole list.

PRO

Summaries

2-sentence summaries for any paper, repo, or model.

Relevance Scoring

The most impactful research lands in your feed and inbox. Papers are picked carefully using our scoring algorithm.

Semantic Search

Search the entire research literature in plain language β€” find papers by meaning, not just keywords, then bookmark them or add to a reading list.

PRO

Related Papers

One click on any paper surfaces the most similar work β€” follow a thread of connected research without starting a new search.

Reading Lists

Group papers into named, color-coded collections like β€œMy PhD reading list” β€” then make any list public to get a beautiful shareable page.

Reading Workflow

Mark any saved paper To read, Reading, or Read, jot private notes, and filter your library by status β€” turn a pile of bookmarks into a workflow you finish.

PRO

Bookmarks & Notes

Save papers, repos, and models. Add personal notes and reorder by drag-and-drop.

PRO

GitHub & HuggingFace

Trending repos and models personalized to your topics. Code badges on papers.

Conference Deadlines

Live countdowns for 100+ AI conferences across ML, vision, NLP, medical AI, robotics, and more. Updated daily from community sources.

Summer Schools & Events

40+ recurring summer & winter schools (MLSS, ICVSS, EEML, MedICSS, RLSS) plus live MICCAI-endorsed event deadlines.

PRO

Email Digest

Get your personalized feed delivered to your inbox daily or weekly. No need to remember to check.

PRO

Paper Q&A

Ask any paper a question and get precise answers grounded in its full text. Methodology, ablations, comparisons, limitations β€” answered in seconds.

PRO

MyResearch

Your personal research assistant. Ask questions across your bookmarks, lists, library, or any specific papers you hand-pick β€” compare papers, find connections, surface what's relevant, and draft related-work with citations.

PRO

Literature Review

Turn any reading list, your bookmarks, or your uploaded library into a structured, cited literature-review draft β€” themes, methods, findings and gaps, synthesized across the papers.

Collaborative Lists

Invite labmates or co-authors to a shared reading list β€” by link or @handle, with editor and viewer roles. Everyone curates the same pile, and you can point MyResearch at it together.

Export to BibTeX

Download any reading list as a .bib file β€” your curated bibliography drops straight into your reference manager or LaTeX editor.

PRO

Save to Zotero

Send any paper to your Zotero library in one click β€” title, authors, abstract, and link mapped cleanly into your reference manager.

Research Community

Connect with other researchers, share findings, and discuss papers with peers. Free for everyone.

Community

Not just papers. A research community.

Post findings, react and discuss, @mention peers, and message researchers directly β€” with live presence so you see who’s online right now.

Online now (8)
MOLS
M
Maya ChenPRO2m
New diffusion sampler β€” 2Γ— faster, no quality loss. Curious what @Omar Farouk thinks for video.
πŸ”₯ 4πŸš€ 2πŸ‘ 6
O
Omar Farouk
Online
Saw your post β€” does it hold temporal consistency?
Yes! Sending the preprint now.
πŸ“„ sampler-v2.pdf

Pricing

Start free. Upgrade when you need more.

Free
$0 forever
  • βœ“3 papers/day
  • βœ“1 topic filter
  • βœ“Reading lists β€” shareable
  • βœ“5 AI semantic searches/mo
  • βœ“Daily AI News (general headlines)
  • βœ“Conference deadlines
  • βœ“Summer schools & events
  • βœ“Community β€” posts, DMs & @mentions
Recommended
Pro50% OFF
$9.99
$4.99/month

Founding Member β€” locked in forever

Less than a cappuccino β˜• for your daily research feed.

Only 487 of 500 founding spots remaining
  • βœ“Your top 5–50 papers daily, picked carefully using our scoring algorithm
  • βœ“2-sentence summaries
  • βœ“All 86 topic filters + All Topics
  • βœ“Custom topics β€” track anything specific (Pro)
  • βœ“Research jobs β€” coming soon
  • βœ“Follow researchers β€” alerts when they publish (in-app + email)
  • βœ“Your Daily AI News β€” personalized, choose 3–15 headlines
  • βœ“GitHub & HuggingFace repos, personalized
  • βœ“Bookmarks with personal notes
  • βœ“Email digest delivered daily or weekly
  • βœ“Paper Q&A (50 questions/month)
  • βœ“MyResearch β€” ask across your bookmarks (15 questions/month)
  • βœ“Semantic search across the literature (100/mo)
  • βœ“Upload your own PDFs β€” a private library with chat & summaries
  • βœ“One-click Save to Zotero
  • βœ“Community + a PRO badge on your profile
$4.99/mo Β· cancel anytime

FAQ

AI research, distilled.

Your personalized daily AI research feed β€” papers, code, and models from every top venue.

Get in touch

Question, feedback, or partnership inquiry? We'd love to hear from you.

Your email
Message