Ax Jun Liu, Zhenglun Kong, Peiyan Dong, Changdi Yang, Tianqi Li, Hao Tang, Geng Yuan, Wei Niu, Wenbin Zhang, Pu Zhao, Xue Lin, Dong Huang, Yanzhi Wang 3/13/2026

Structured Agent Distillation for Large Language Model

Framework for compressing large LLM-based ReAct agents into smaller student models while preserving reasoning and action consistency.

Ax Matthieu Blanke, Yongquan Qu, Sara Shamekh, Pierre Gentine 3/13/2026

Strictly Constrained Generative Modeling via Split Augmented Langevin Sampling

Framework for generative modeling with enforced physical constraints using split augmented Langevin sampling for scientific applications.

Ax Mengjie Zhao, Olga Fink 3/13/2026

Disentangling Slow and Fast Temporal Dynamics in Degradation Inference with Hierarchical Differential Models

Hierarchical differential model for inferring system degradation from sensor data by disentangling slow and fast temporal dynamics.

Ax Jiajun Bao, Nicolas Boull\'e, Toni J. B. Liu, Rapha\"el Sarfati, Christopher J. Earls 3/13/2026

Text-Trained LLMs Can Zero-Shot Extrapolate PDE Dynamics, Revealing a Three-Stage In-Context Learning Mechanism

Text-trained LLMs perform zero-shot extrapolation of PDE dynamics, revealing three-stage in-context learning mechanism for spatiotemporal forecasting.

Ax Cl\'ement Bonet, Elsa Cazelles, Lucas Drumetz, Nicolas Courty 3/13/2026

Busemann Functions in the Wasserstein Space: Existence, Closed-Forms, and Applications to Slicing

Mathematical study of Busemann functions in Wasserstein space with applications to geometric machine learning and data slicing.

Ax Ozgur Guldogan, Neeraj Sarna, Yuanyuan Li, Michael Berger 3/13/2026

Counterfactually Fair Conformal Prediction

Development of conformal prediction method that ensures counterfactual fairness in prediction sets for fair decision-making under uncertainty.

Ax Wanhao Yu, Zheng Wang, Shuteng Niu, Sen Lin, Li Yang 3/13/2026

More Than Memory Savings: Zeroth-Order Optimization Mitigates Forgetting in Continual Learning

Zeroth-order optimization approach for continual learning that improves memory efficiency and addresses plasticity-stability tradeoffs without gradient computation.

Ax Eric Bigelow, Daniel Wurgaft, YingQiao Wang, Noah Goodman, Tomer Ullman, Hidenori Tanaka, Ekdeep Singh Lubana 3/13/2026

Belief Dynamics Reveal the Dual Nature of In-Context Learning and Activation Steering

Research on unifying in-context learning and activation steering as instances of a broader framework using belief dynamics to control LLM behavior at inference time.

Ax Xiuying Wei, Caglar Gulcehre 3/13/2026

RAT+: Train Dense, Infer Sparse -- Recurrence Augmented Attention for Dilated Inference

Introduces RAT+, structured dilated attention architecture that enables sparse inference while maintaining long-range connectivity and accuracy.

Ax Zhuoxu Huang, Mengxi Jia, Hao Sun, Xuelong Li, Jungong Han 3/13/2026

Controllable Exploration in Hybrid-Policy RLVR for Multi-Modal Reasoning

Proposes controllable exploration strategy for RLVR training of multi-modal LLMs to address entropy collapse and policy degradation.

Ax Jose Javier Gonzalez Ortiz, Abhay Gupta, Christopher Rinard, Davis Blalock 3/13/2026

FlashOptim: Optimizers for Memory-Efficient Training

Introduces FlashOptim, memory-efficient optimizers for mixed-precision neural network training reducing per-parameter memory requirements.

Ax Isotta Magistrali, Fr\'ed\'eric Berdoz, Sam Dauncey, Roger Wattenhofer 3/13/2026

Subliminal Signals in Preference Labels

Shows preference labels in LLM-as-judge training can function as covert communication channels, challenging assumptions about semantic supervision.

Ax Hadi Sotoudeh, Payel Mukhopadhyay, Ruben Ohana, Michael McCabe, Neil D. Lawrence, Shirley Ho, Miles Cranmer 3/13/2026

On the Value of Tokeniser Pretraining in Physics Foundation Models

Investigates tokenizer pretraining impact on physics foundation models for emulating complex multiphysics phenomena in data-limited settings.

Ax Joohyung Lee, Kwanhyung Lee, Changhun Kim, Eunho Yang 3/13/2026

Structure-Aware Set Transformers: Temporal and Variable-Type Attention Biases for Asynchronous Clinical Time Series

Structure-aware set transformers with temporal and variable-type attention for asynchronous clinical time series in EHR data.

Ax Tatjana Krau, Jorge Mandlmaier, Tobias Damm, Frieder Heieck 3/13/2026

Impact of Markov Decision Process Design on Sim-to-Real Reinforcement Learning

Analyzes how MDP design choices (state composition, rewards, dynamics) affect sim-to-real transfer in reinforcement learning for industrial control.

Ax Kyungryeol Lee, Kyeonghyun Lee, Seongmin Hong, Byung Hyun Lee, Se Young Chun 3/13/2026

Unlearning the Unpromptable: Prompt-free Instance Unlearning in Diffusion Models

Instance unlearning method for diffusion models removing specific outputs without text prompts, addressing unpromptable undesired generations.

Ax Zegu Zhang, Jian Zhang 3/13/2026

Historical Consensus: Preventing Posterior Collapse via Iterative Selection of Gaussian Mixture Priors

Proposes iterative selection of Gaussian mixture priors to prevent posterior collapse in variational autoencoders.

Ax Benjamin A. T. Grahama, Lauren Brown, Georgios Chochlakis, Morteza Dehghani, Raquel Delerme, Brittany Friedman, Ellie Graeden, Preni Golazizian, Rajat Hebbar, Parsa Hejabi, Aditya Kommineni, Mayag\"uez Salinas, Michael Sierra-Ar\'evalo, Jackson Trager, Nicholas Weller, Shrikanth Narayanan 3/13/2026

Community-Informed AI Models for Police Accountability

AI system analyzing police bodycam footage at scale to assess officer-public interactions and improve government accountability.

Ax Han Qi, Haocheng Yin, Aris Zhu, Yilun Du, Heng Yang 3/13/2026

Inference-Time Enhancement of Generative Robot Policies via Predictive World Modeling

Generative Predictive Control method augments frozen diffusion policies with action-conditioned world models for improved robot control without retraining.

Ax Liam Welsh, Udit Grover, Sebastian Jaimungal 3/13/2026

Multi-Agent Reinforcement Learning for Greenhouse Gas Offset Credit Markets

Applies multi-agent reinforcement learning to greenhouse gas offset credit markets for emissions control and carbon project trading simulation.

Ax Aida Kostikova, Zhipin Wang, Deidamea Bajri, Ole P\"utz, Benjamin Paa{\ss}en, Steffen Eger 3/13/2026

LLLMs: A Data-Driven Survey of Evolving Research on Limitations of Large Language Models

Data-driven survey identifying 14,648 papers on LLM limitations from 2022-2025 using automated classification and expert validation across 250,000 academic papers.

Ax Tom Danino, Nahum Shimkin 3/13/2026

Enhancing Sample Efficiency in Multi-Agent RL with Uncertainty Quantification and Selective Exploration

Novel algorithm for multi-agent reinforcement learning using uncertainty quantification and selective exploration to improve sample efficiency in joint action spaces.

Ax Nadav Kunievsky, James A. Evans 3/13/2026

Measuring Intent Comprehension in LLMs

Research paper on measuring whether LLMs comprehend user intent beyond surface-level text patterns, addressing training-inference gaps in language models.

Ax Peibo Li, Shuang Ao, Hao Xue, Yang Song, Maarten de Rijke, Johan Barth\'elemy, Tomasz Bednarz, Flora D. Salim 3/13/2026

Refine-POI: Reinforcement Fine-Tuned Large Language Models for Next Point-of-Interest Recommendation

Reinforcement fine-tuning approach for LLMs applied to point-of-interest recommendation with improved semantic indexing.

Ax Orion Weller, Kathryn Ricci, Marc Marone, Antoine Chaffin, Dawn Lawrie, Benjamin Van Durme 3/13/2026

Seq vs Seq: An Open Suite of Paired Encoders and Decoders

Open benchmark suite comparing paired encoder and decoder architectures for NLP tasks with controlled parameter counts.

Ax Ondrej Bohdal, Mete Ozay, Jijoong Moon, Kyeng-Hun Lee, Hyeonmok Ko, Umberto Michieli 3/13/2026

Efficient Compositional Multi-tasking for On-device Large Language Models

Adapter parameters for efficient multi-task LLM inference on-device via task merging for compositional learning.

Ax Sayan Nag, K J Joseph, Koustava Goswami, Vlad I Morariu, Balaji Vasan Srinivasan 3/13/2026

Agentic Design Review System

Agentic Design Review System orchestrates multiple AI agents to collaboratively analyze graphic designs with meta-agent coordination.

Ax Orion Weller, Michael Boratko, Iftekhar Naim, Jinhyuk Lee 3/13/2026

On the Theoretical Limitations of Embedding-Based Retrieval

arXiv paper analyzes theoretical limitations of embedding-based retrieval for diverse tasks including reasoning and code generation.

Ax Haoyang Zheng, Xinyang Liu, Cindy Xiangrui Kong, Nan Jiang, Zheyuan Hu, Weijian Luo, Wei Deng, Guang Lin 3/13/2026

Ultra-Fast Language Generation via Discrete Diffusion Divergence Instruct

DiDi-Instruct trains fast few-step language generation via distillation from discrete diffusion LLMs while maintaining quality.

Ax Henrique Assump\c{c}\~ao, Diego Ferreira, Leandro Campos, Fabricio Murai 3/13/2026

CodeEvolve: an open source evolutionary coding agent for algorithmic discovery and optimization

CodeEvolve open-source framework combines LLMs with evolutionary algorithms to synthesize optimized algorithmic solutions guided by execution feedback.

Ax Atsuyuki Miyai, Mashiro Toyooka, Takashi Otonari, Zaiying Zhao, Kiyoharu Aizawa 3/13/2026

Jr. AI Scientist and Its Risk Report: Autonomous Scientific Exploration from a Baseline Paper

Jr. AI Scientist is an autonomous research system that mimics novice researcher workflows, conducting autonomous exploration and including risk analysis.

Ax Andr\'es Carofilis, Sergio Burdisso, Esa\'u Villatoro-Tello, Shashi Kumar, Kadri Hacioglu, Srikanth Madikeri, Pradeep Rangappa, Manjunath K E, Petr Motlicek, Shankar Venkatesan, Andreas Stolcke 3/13/2026

Text-only adaptation in LLM-based ASR through text denoising

Novel text-only adaptation method for LLM-based ASR systems using text denoising to preserve speech-text alignment without fine-tuning disruption.

Ax Zelai Xu, Zhexuan Xu, Ruize Zhang, Chunyang Zhu, Shi Yu, Weilin Liu, Quanlu Zhang, Wenbo Ding, Chao Yu, Yu Wang 3/13/2026

WideSeek-R1: Exploring Width Scaling for Broad Information Seeking via Multi-Agent Reinforcement Learning

Multi-agent reinforcement learning system for width-scaled information seeking, exploring complementary depth scaling in LLM deployment.

Ax Xinwu Ye, Yicheng Mao, Jia Zhang, Yimeng Liu, Li Hao, Fang Wu, Zhiwei Li, Yuxuan Liao, Zehong Wang, Zhiyuan Liu, Zhenfei Yin, Li Yuan, Philip Torr, Huan Sun, Xiangxiang Zeng, Mengdi Wang, Le Cong, Shenghua Gao, Xiangru Tang 3/13/2026

LatentChem: From Textual CoT to Latent Thinking in Chemical Reasoning

LatentChem: Latent reasoning interface for chemical LLMs, decoupling chemical reasoning from text tokens for improved efficiency.

Ax Kwanghee Choi, Eunjung Yeo, Cheol Jun Cho, David Harwath, David R. Mortensen 3/13/2026

[b]=[d]-[t]+[p]: Self-supervised Speech Models Discover Phonological Vector Arithmetic

Study of phonological vector arithmetic in self-supervised speech models across 96 languages, analyzing representation structure.

Ax Rafael R. Baptista, Andr\'e de Lima Salgado, Ricardo V. Godoy, Marcelo Becker, Thiago Boaventura, Gustavo J. G. Lahr 3/13/2026

Evaluating Zero-Shot and One-Shot Adaptation of Small Language Models in Leader-Follower Interaction

Evaluation of small language models for role classification in human-robot interaction with zero-shot and one-shot adaptation.

Ax Brandon Yee, Lucas Wang, Kundana Kommini, Krishna Sharma 3/13/2026

Geodesic Semantic Search: Learning Local Riemannian Metrics for Citation Graph Retrieval

Retrieval system learning node-specific Riemannian metrics on citation graphs for geometry-aware semantic search.

Ax Fr\'ed\'eric Berdoz, Leonardo Rugli, Roger Wattenhofer 3/13/2026

Can AI Agents Agree?

Systematic evaluation of LLM-based AI agents in Byzantine consensus games, testing agreement behavior in adversarial settings.

Ax Fr\'ed\'eric Berdoz, Yann Billeter, Yann Vonlanthen, Roger Wattenhofer 3/13/2026

Reasoning Boosts Opinion Alignment in LLMs

Study of reasoning techniques in LLMs for political opinion modeling and alignment with individual preferences.

Ax Siddharth Boppana, Annabel Ma, Max Loeffler, Raphael Sarfati, Eric Bigelow, Atticus Geiger, Owen Lewis, Jack Merullo 3/13/2026

Reasoning Theater: Disentangling Model Beliefs from Chain-of-Thought

Analysis of chain-of-thought reasoning in LLMs, comparing activation probing and early stopping across DeepSeek-R1 and GPT-OSS models.

Ax Louis Mozart Kamdem Teyou, Caglar Demir, Axel-Cyrille Ngonga Ngomo 3/13/2026

Semantics-Aware Caching for Concept Learning

Caching optimization for concept learning in description logic knowledge bases using supervised learning.

Ax Vinzenz Thoma, Georgios Piliouras, Luke Marris 3/13/2026

Deep Incentive Design with Differentiable Equilibrium Blocks

Differentiable equilibrium blocks for multi-agent incentive design in game theory and economics applications.

Ax Yuzhi Liu, Duo Zhang, Anyang Peng, Weinan E, Linfeng Zhang, Han Wang 3/13/2026

Scaling Machine Learning Interatomic Potentials with Mixtures of Experts

Mixture-of-Experts architectures for machine learning interatomic potentials with analysis of routing strategies and sparse activation.

Ax Yuling Jiao, Yanming Lai, Huazhen Lin, Wensen Ma, Houduo Qi, Defeng Sun 3/13/2026

Beyond the Prompt in Large Language Models: Comprehension, In-Context Learning, and Chain-of-Thought

Research into theoretical mechanisms of LLM phenomena: semantic prompt comprehension, in-context learning, and chain-of-thought reasoning.

Ax Yannis Karmim (ALMAnaCH), Renato Pino (UCHILE), Hernan Contreras (UCHILE), Hernan Lira (CENIA), Sebastian Cifuentes (CENIA), Simon Escoffier (PUC), Luis Mart\'i (ALMAnaCH), Djam\'e Seddah (ALMAnaCH), Valentin Barri\`ere (UCHILE, CENIA) 3/13/2026

Leveraging Wikidata for Geographically Informed Sociocultural Bias Dataset Creation: Application to Latin America

Dataset creation using Wikidata to detect sociocultural biases in LLMs, focusing on Latin American languages and cultures.

HN daniel_iversen 3/13/2026

Fast non-Chromium browser for AI agents: LightPanda

LightPanda is a headless browser built for AI agent workloads with optimized performance, security, and CDP support.

HN andes314 3/13/2026

Give Your AI Agent a Live Status Page

ClawJetty adds live status pages and progress tracking to AI agents with shareable run links and task completion visibility.

HN opcastil11 3/13/2026

Prowl – An agent discovery network (ASO for AI agents)

Prowl discovers and ranks AI services/APIs via natural language queries. Claude tests services across 8 performance dimensions.

HN mooreds 3/13/2026

Agent harness for building analytics into your app on top of ClickHouse

Developer framework enabling AI agents to build analytics applications on ClickHouse. Provides specialized interfaces for coding agents.

HN gmays 3/13/2026

Teaching LLMs to reason like Bayesians

Research from Google on training LLMs to reason using Bayesian inference for improved probabilistic world modeling in agentic systems.