| tech.blog / monetary.live / ivan.digital
May 05, 2026 at 07:11 UTC
🤗 HuggingFace Daily Papers
From Context to Skills: Can Language Models Learn from Context Skillfully?
▲ 20 🎓 Tsinghua University
  • A self-evolving framework autonomously discovers and refines context-specific skills for language models through a multi-agent self-play loop with Cha
MolmoAct2: Action Reasoning Models for Real-world Deployment
▲ 13 🎓 Ai2
  • MolmoAct2 presents an open-action reasoning model for robotics that improves upon previous systems through specialized vision-language-model backbones
Generative Modeling with Orbit-Space Particle Flow Matching
▲ 1 🎓 Georgia Institute of Technology
  • Orbit-Space Geometric Probability Paths (OGPP) presents a particle-native flow-matching framework that improves generative modeling of particle system
PhysicianBench: Evaluating LLM Agents in Real-World EHR Environments
▲ 1 🎓 Stanford University
  • PhysicianBench evaluates LLM agents on real clinical tasks requiring complex, multi-step workflows within electronic health record environments, revea
Perceptual Flow Network for Visually Grounded Reasoning
▲ 0 🎓 ECNU
  • Perceptual Flow Network addresses limitations in vision-language models by decoupling perception from reasoning and using variational reinforcement le
Context Learning 1Language Models 1Skill Augmentation 1Multi-Agent Self-Play 1Challenger 1Reasoner 1Judge 1Skill Refinement 1Adversarial Collapse 1Cross-Time Replay 1Cl-Bench 1Vision-Language-Action Models 1
🏛️ Top Research Institutions
HalluScan: A Systematic Benchmark for Detecting and Mitigating Hallucinations in Instructi
🏛️ Sofrecom AI & Machine Learning
  • Large Language Models (LLMs) often generate inaccurate or misleading content, known as hallucinations.
  • HalluScan provides a systematic benchmark for detecting and mitigating these hallucinations in instruction-following LLM
Affinity Tailor: Dynamic Locality-Aware Scheduling at Scale
🏛️ Google Systems & Infrastructure
  • Modern large multicore systems struggle with efficient CPU scheduling when multiple workloads share resources.
  • A dynamic locality-aware scheduling algorithm that optimizes CPU usage by considering workload affinities.
Testing, Credible Compilation, and Verification in the Axon Verified Compiler in Lean and
🏛️ National University of Singapore Software & Programming
  • Ensuring the correctness of compiled code is critical but challenging due to the complexity of modern programming langua
  • The Axon compiler integrates testing, credible compilation, and verification with machine-checked proofs to guarantee co
Exponential speedups in fault-tolerant processing of quantum experiments
🏛️ Harvard Quantum Initiative Theory & Algorithms
  • Quantum information processing is hindered by noise introduced when coupling quantum processors to experimental samples.
  • The paper presents methods for achieving exponential speedups in fault-tolerant processing of quantum experiments.
Greed for the Spheres: A Signed Distance Interpolation Method
🏛️ University of Southern California Applications
  • Interpolating Signed Distance Function (SDF) data from discrete samples is challenging and often inconsistent.
  • A new method that ensures the interpolated SDF data values maintain full consistency with the original samples.
A Levered ETF Anomaly Explained
🏛️ University of California, Berkeley Quantitative Finance
  • Leveraged ETFs often exhibit counterintuitive performance relative to their underlying indices, leading to investor conf
  • An analysis explaining the anomaly of leveraged ETFs in relation to the S&P 500 Index's performance.
AI & Machine Learning
414 papers 7 cats
Systems & Infrastructure
12 papers 6 cats
Software & Programming
33 papers 4 cats
Theory & Algorithms
26 papers 5 cats
Applications
102 papers 7 cats
Other CS
18 papers 8 cats
Quantitative Finance
12 papers 9 cats
Economics
10 papers 3 cats