HN AlatAnmoku 5/25/2026

Show HN: Porting my Newsletter to MCP – You set WHEN and HOW OFTEN to receive it

Newsletter platform ported to Model Context Protocol allowing subscribers to customize delivery time and frequency.

HN loaderchips 5/25/2026

Show HN: YieldOS-Lite – A simulator for LLM inference control-plane governance

Simulator for LLM inference control-plane governance exploring heterogeneous workload scheduling vs mechanistic schedulers like continuous batching.

Ax Joss Armstrong 5/25/2026

BOHM: Zero-Cost Hierarchical Attribution for Compound AI Systems

BOHM: zero-cost hierarchical attribution method for compound AI systems decomposing component contributions without requiring arbitrary subset evaluation.

Ax Zelin Zhao, Bo Yuan, Jaemoo Choi, Yongxin Chen 5/25/2026

RMA: an Agentic System for Research-Level Mathematical Problems

Research Math Agents (RMA): agentic framework for automated reasoning on research-level mathematical problems with long-horizon proof refinement.

Ax Shuofei Qiao, Yunxiang Wei, Jiazheng Fan, Bin Wu, Busheng Zhang, Mengru Wang, Yuqi Zhu, Ningyu Zhang, Keyan Ding, Qiang Zhang, Huajun Chen 5/25/2026

SciAtlas: A Large-Scale Knowledge Graph for Automated Scientific Research

SciAtlas: large-scale knowledge graph for scientific research enabling topological reasoning and interdisciplinary integration for AI agents.

Ax Deepak Panigrahy, Aakash Tyagi 5/25/2026

Energy per Successful Goal: Goal-Level Energy Accounting for Agentic AI Systems

Energy accounting framework for agentic AI systems measuring energy consumption per completed goal rather than per inference invocation.

Ax Riyaz Ahuja, Tate Rowney, Jeremy Avigad, Sean Welleck 5/25/2026

ImProver 2: Iteratively Self-Improving LMs for Neurosymbolic Proof Optimization

ImProver 2: neurosymbolic framework for automated optimization of formal mathematical proofs using iteratively self-improving language models.

Ax Bruno Deprez, Wouter Verbeke, Tim Verdonck 5/25/2026

Is TabPFN the Silver Bullet for Insurance Pricing?

Evaluation of TabPFN tabular foundation model versus traditional methods for insurance pricing and claim modeling tasks.

Ax Jingyan Zhang, Han Liang, Ruichi Zhang, Bin Li, Juze Zhang, Xin Chen, Jingya Wang, Lan Xu, Jingyi Yu 5/25/2026

SCRIPT: Scalable Diffusion Policy with Multi-stage Training for Language-driven Physics-Based Humanoid Control

SCRIPT: scalable diffusion policy with multi-stage training for language-driven control of physics-based humanoid embodied agents.

Ax Ruofan Jin, Zaixi Zhang 5/25/2026

Agentic-VLA: Efficient Online Adaptation for Vision-Language-Action Models

Agentic-VLA framework enabling efficient online adaptation of vision-language-action models for robotic manipulation tasks.

Ax Shubham Parashar, Atharv Chagi, Jacob Helwig, Lakshmi Jotsna, Sushil Vemuri, James Caverlee, Dileep Kalathil, Shuiwang Ji 5/25/2026

Learnability-Informed Fine-Tuning of Diffusion Language Models

Novel fine-tuning approach for diffusion language models that accounts for token learnability to improve reasoning capabilities.

Ax Benedikt L\"utke Schwienhorst, Nadja Klein, Johannes Lederer 5/25/2026

Diffusion-based Denoising Beats Vanilla Score Matching in Parameter Estimation: A Theoretical Explanation

Theoretical comparison of diffusion-based denoising score matching versus vanilla score matching for parameter estimation in multimodal distributions.

Ax Tobias von Arx, Tanguy Dieudonn\'e 5/25/2026

Memorization Dynamics of Fill-in-the-Middle Pretraining

Study on verbatim memorization in fill-in-the-middle pretraining for causal language models, comparing with left-to-right objectives.

Ax Helena Merker, Nick Walker, Andreea Bobu 5/25/2026

Robots That Know What to Ask: Recovering Misaligned Rewards through Targeted Explanations

Method for robots to query humans and recover misaligned reward functions through targeted explanations when demonstrations lack coverage.

Ax Yurii Kvasiuk, Tianyi Li, Owen Colegrove, Moritz M\"unchmeyer 5/25/2026

MadEvolve: Evolutionary Optimization of Trading Systems with Large Language Models

LLM-driven evolutionary algorithm optimization framework for quantitative finance and Bitcoin trading strategy generation.

Ax Dongxin Guo 5/25/2026

The Deterministic Horizon: Impossibility Results as Design Specifications for Trustworthy AI Systems

Framework converting impossibility results into design specifications for trustworthy AI, proving architecture-based accuracy ceilings for LLMs.

Ax Dongxin Guo, Jikun Wu, Siu Ming Yiu 5/25/2026

Do Language Models Know What Not to Say? Causal Evidence for Statistical Preemption in LLMs

Computational study on statistical preemption in LLMs, examining how models learn acceptable linguistic forms without negative evidence.

Ax Mantas Skackauskas, Xinyue Hao, Laura Sevilla-Lara 5/25/2026

The TIME Machine: On The Power of Motion for Efficient Perception

Video representation learning framework improving efficiency and reducing scaling costs while learning beyond language-constrained approaches.

Ax Yury Belousov, Brian Pulfer, Vitaliy Kinakh, Slava Voloshynovskiy 5/25/2026

Dithering Defense: Adversarial Robustness of Vision Foundation Models via Multi-Level Floyd-Steinberg Dithering

Adversarial robustness technique using multi-level Floyd-Steinberg dithering as lightweight input transformation for vision foundation models.

Ax Colin Gaffney, Shutong Li, Daniel Ng, Anastasia Petrushkina, Niket Kumar, Adam Cogdell, Mridul Sahu, Yaning Liang, Nikhil Bansal, Justin Pan, Angel Mau, Abhishek Agrawal, Marco Berlot, Ruoxin Sang, Kiranbir Sodhia, Rakesh Iyer 5/25/2026

Orbax: Distributed Checkpointing with JAX

Orbax: JAX-native distributed checkpointing library abstracting complexity of accelerator systems for ML model lifecycle management.

Ax Stelios Boulitsakis Logothetis, Angela Wood, Pietro Li \`o 5/25/2026

KAPLAN: Kolmogorov-Arnold Prognostic Learnable Activation Networks for Survival Analysis

Kolmogorov-Arnold Network variant for nonparametric survival analysis in clinical settings, replacing manual feature engineering.

Ax Philipp Kern, Lorenzo Rovida, Samuel Teuber, Edoardo Manino, Carsten Sinz, Alberto Leporati 5/25/2026

Encrypted Neural Networks without Overflows

Methods for evaluating neural networks on encrypted data using fully homomorphic encryption without overflow issues in CKKS scheme.

Ax Caleb Skinner, Yihan Guo, Meng Li 5/25/2026

LLM Sparsity Prior for Robust Feature Selection

Framework for using LLMs to generate domain-informed priors for feature selection, with methods to quantify weight quality and improve robustness.

Ax Mingyuan Fan, Yu Liu, Fuyi Wang, Cen Chen 5/25/2026

What Does the Server See? Understanding Privacy Leakage from Large Language Models in Split Inference

Research on privacy leakage in split inference for LLMs deployed on resource-constrained devices, examining what intermediate activations reveal to servers.

Ax Luze Sun, Anshuman Suri, Harsh Chaudhari, Cristina Nita-Rotaru, Alina Oprea 5/25/2026

PoisonForge: Task-Level Targeted Poisoning Benchmark for Instruction-Tuned LLMs

Benchmark for task-level data poisoning attacks on instruction-tuned LLMs, parameterizing threat model across multiple dimensions.

Ax Chuyifei Zhang, Hongyu Cui, Xiaowen Huang, Jitao Sang 5/25/2026

Positional Failures in Long-Context LLMs: A Blind Spot in Reasoning Benchmarks

Audit of long-context LLM benchmarks revealing that position-controlled evaluation is missing from mainstream reasoning benchmarks.

Ax Baturay Saglam, Dionysis Kalogerias 5/25/2026

Self-Improving In-Context Learning

Method for optimizing few-shot prompt embeddings at test time using log-probabilities signal to improve in-context learning performance.

Ax Vartan Shadarevian, Kia Ghods, Alex Kenich, Anany Kotawala 5/25/2026

GENSTRAT: Toward a Science of Strategic Reasoning in Large Language Models

Framework and benchmark for evaluating strategic reasoning capabilities in LLMs deployed as economic agents in auctions and markets.

Ax Jiahao Shi, Omar Hagrass, Jason M. Klusowski 5/25/2026

Coupled Training with Privileged Information and Unlabeled Data

Training method using privileged information and unlabeled data via coupled models for improved deployment-time prediction.

Ax Zehao Wang, Shilong Jin, Zhao Cao, Lanjun Wang 5/25/2026

When Planning Fails Despite Correct Execution: On Epistemic Calibration for LLM-Based Multi-Agent Systems

Analysis of epistemic miscalibration in LLM-based multi-agent systems where agents misjudge knowledge despite correct execution.

Ax Mario Markov (INSAIT, Sofia University "St. Kliment Ohridski"), Stefan Maria Ailuro (INSAIT, Sofia University "St. Kliment Ohridski"), Mohammad Mahdi (INSAIT, Sofia University "St. Kliment Ohridski"), Luc Van Gool (INSAIT, Sofia University "St. Kliment Ohridski"), Danda Pani Paudel (INSAIT, Sofia University "St. Kliment Ohridski") 5/25/2026

B-GRTO: Bootstrapped Group Relative Tool Optimization for Referring Segmentation

Bootstrapped group relative tool optimization for referring segmentation using vision-language models with reinforcement learning.

Ax Vipul Gupta, Shikhar Mohan, Lakshya Kumar, Pranjal Chitale, Nikit Begwani, Amit Singh, Manik Varma 5/25/2026

HARNESS-LM: A Three-Phase Training Recipe for Harnessing SLMs in Sponsored Search Retrieval

Three-phase training framework for deploying small language models in high-throughput sponsored search retrieval with reduced latency.

Ax John Sous, Michael Winer 5/25/2026

Asymmetric Scaling Laws from Sparse Features

Theoretical model of neural scaling laws under sparse feature activations, analyzing loss behavior in under/overparameterized regimes.

Ax Shivam Chourasia, Hitesh Kapoor, Nilesh Patil 5/25/2026

Structure-Guided Entity Resolution: Fine-Tuning LLMs for Robust Name Matching in Complex Linguistic Contexts

Fine-tuning LLMs for entity resolution and name matching in complex linguistic contexts, applied to KYC compliance.

Ax Matthias Cosler, Cas Cremers, Bernd Finkbeiner, Mohamed Ghanem, Niklas Medinger 5/25/2026

Less Effort, Shorter Proofs: Reinforcement Learning for Security Protocol Analysis in Tamarin

RL-based framework inspired by AlphaZero for automating security protocol analysis in Tamarin tool.

Ax Nikola Pavlovic, Sattar Vakili, Qing Zhao 5/25/2026

Learning Kernel-Based MDPs from Episodic Preferential Feedback

Theoretical framework for learning kernel-based MDPs from preferential feedback in episodic reinforcement learning.

Ax Liupeng Li, Haoqian Kang, Zhenyu Lu, Jinpeng Wang, Bin Chen, Ke Chen, Yaowei Wang 5/25/2026

CVSearch: Empowering Multimodal LLMs with Cognitive Visual Search for High-Resolution Image Perception

CVSearch method enhancing multimodal LLMs for high-resolution image perception through cognitive visual search strategies.

Ax Qian Zhang, George Em Karniadakis 5/25/2026

Operator Learning for Reconstructing Flow Fields from Sparse Measurements: a Language Model Approach

Operator learning framework using language model architecture for reconstructing flow fields from sparse measurements.

Ax Seyed Bagher Hashemi Natanzi, Pranshav Gajja, Bo Tang, Vijay K. Shah 5/25/2026

Advanced AI Service Provisioning in O-RAN through LLM Engine Integration

LLM engine integration with O-RAN for AI service provisioning, automating xApp/rApp creation, data collection, and code generation.

Ax Luke Braithwaite, Alessio Borgi, Gabriele Onorato, Kristjan Tarantelli, Francesco Restuccia, Fabrizio Silvestri, Pietro Li\`o 5/25/2026

Heterogeneous Sheaf Neural Networks

Heterogeneous graph neural networks for handling nodes/edges with different types and feature spaces across domains.

Ax Aecheon Jung, Seunghwan Lee, Dongyoon Han, Sungeun Hong 5/25/2026

SyMerge: From Non-Interference to Synergistic Merging via Single-Layer Adaptation

Model merging technique that enables task synergy by adapting single layers to improve cross-task performance compatibility.

Ax Peiliang Zhang, Jingling Yuan, Qing Xie, Yongjun Zhu, Lin Li 5/25/2026

Representational Alignment with Chemical Induced Fit for Molecular Relational Learning

Molecular relational learning with chemical-informed representation alignment for predicting binding site compatibility.

Ax Christian Kl\"otergens, Vijaya Krishna Yalavarthi, Randolf Scholz, Maximilian Stubbemann, Stefan Born, Lars Schmidt-Thieme 5/25/2026

Physiome-ODE: A Benchmark for Irregularly Sampled Multivariate Time Series Forecasting Based on Biological ODEs

Benchmark dataset for irregularly sampled multivariate time series forecasting based on biological ODE systems.

Ax Thong Pham, Takashi Nicholas Maeda, Shohei Shimizu 5/25/2026

Causal Additive Models with Unobserved Causal Paths and Backdoor Paths

Causal discovery framework identifying causal directions with unobserved backdoor/causal paths in additive models.

Ax Zhong Li, Qi Huang, Lincen Yang, Jiayang Shi, Zhao Yang, Niki van Stein, Thomas B\"ack, Matthijs van Leeuwen 5/25/2026

Diffusion and Flow Matching Models for Tabular Data: A Survey

Survey of diffusion and flow matching models for tabular data generation, addressing missing values and domain constraints.

Ax Keisuke Kawano, Takuro Kutsuna, Naoki Hayashi, Yasushi Esaki, Hidenori Tanaka 5/25/2026

CT-OT Flow: Estimating Continuous-Time Dynamics from Discrete Temporal Snapshots

Optimal transport flow method for estimating continuous-time dynamics from discrete temporal snapshots with uncertainty.

Ax Han Wan, Rui Zhang, Hao Sun 5/25/2026

Spectral-inspired Operator Learning with Limited Data and Unknown Physics

Spectral-inspired neural operator learning PDE dynamics from 2-5 trajectories without explicit physics.

Ax R. Sharma, Y. B. Guo 5/25/2026

Physics-Informed Machine Learning Regulated by Finite Element Analysis for Simulation Acceleration of Melt Pool Dynamics in Laser Powder Bed Fusion

Physics-informed neural networks regulated by FEA for accelerating melt pool dynamics simulation in laser manufacturing.

Ax Joon-Hyun Park, Mujin Cheon, Jeongsu Wi, Dong-Yeun Koh 5/25/2026

BOOST: A Data-Driven Framework for the Automated Joint Selection of Kernel and Acquisition Functions in Bayesian Optimization

Automated data-driven framework for jointly selecting kernel and acquisition functions in Bayesian optimization.

Ax Reijo Jaakkola, Tomi Janhunen, Antti Kuusisto, Magdalena Ortiz, Matias Selin, Mantas \v{S}imkus 5/25/2026

Graph Learning via Logic-Based Weisfeiler-Leman Variants and Tabularization

Graph classification via Weisfeiler-Leman algorithm variants with tabularization and logical framework modifications.