AI summaries of arXiv papers
Paste any arXiv URL or paper ID. Or just swap arxiv.org → arxivtldr.org in your address bar.
Add to Chrome — Free ExtensionToday's Papers
Probing the Reionization History and Bubble Sizes with JWST Lyman-α Fraction Measurements
This paper uses JWST Lyman-alpha fraction measurements to constrain the reionization history and ionized bubble sizes in the early universe.
Multimodal Flow: Unified Flow Modeling of Language and Vision in Embedding Spaces
Multimodal Flow introduces a unified, fully continuous generative model for language and vision, outperforming hybrid and discrete models.
Ranking-Aware Prompt Optimization for Multimodal Clinical Diagnosis
This paper introduces Ranking-PE, a prompt optimization method for MLLMs in clinical diagnosis that optimizes for AUROC, outperforming accuracy-based methods.
Semifactual Credit-Augmented Policy Optimization
This paper introduces SCAPO, a causally inspired RLVR method that uses semifactual stability for finer-grained token-level credit assignment to improve LLM reasoning.
Removing Timing Shortcuts Improves Non-Invasive Brain-to-Text
This paper identifies and removes a timing shortcut in non-invasive brain-to-text decoding, significantly improving performance by focusing on brain activity.
Physis-Lang: Self-Evolving Language as a Physical Representation for Video World Model
Physis-Lang introduces a self-evolving language framework to improve video world models' physical plausibility, outperforming proprietary models.
ViTeX-Bench: Benchmarking High-Fidelity Video Scene Text Editing
ViTeX-Bench introduces a new benchmark and dataset for high-fidelity video scene text editing, evaluating text correctness, visual quality, and temporal consistency.
AssemblyWorld: Rethinking 3D Assembly with General-Purpose Agents
AssemblyWorld is a new 3D environment and benchmark for evaluating general-purpose agents on complex assembly tasks using visual interaction.
On The Simplest Quantum-Secure Block Cipher
This paper proves the two-round Even-Mansour cipher is quantum-secure against adaptive adversaries and shows its minimality.
Image Classifiers are Efficient Self-Supervised Video Representation Learners
VideoMSN uses image Vision Transformers with masked Siamese networks to efficiently learn spatio-temporal video representations.
Ego4WAM: What Matters When Scaling Egocentric Human Data for Robot Learning?
Ego4WAM systematically studies how egocentric human data properties like alignment, diversity, and supervision impact robot learning performance.
EvoDuet: Bilevel Co-Evolution of Web Searching and Task Solving for Scientific Discovery
EvoDuet is a bi-level co-evolution method that optimizes both solutions and web search queries for LLMs to enhance scientific discovery.
Browse by Category
Browse by author →Browse by Topic
All topics →📬 Weekly AI Paper Digest
Get the top 10 AI/ML arXiv papers from the week — summarized, scored, and delivered to your inbox every Monday.
The URL swap trick
Reading a paper on arxiv.org? Just change the domain to arxivtldr.org in your address bar:
Paste a link
Enter any arXiv URL or paper ID
AI summarizes
Get a TLDR, key bullets, and why it matters
Read & share
Get key insights in seconds, share with colleagues