All entries

Policy Library CBF: Finite-Horizon Safety at Runtime via Parallel Rollouts

arXiv:2605.16588v1 Announce Type: new Abstract: Safety-critical autonomy in unstructured environments poses significant challenges for online safety certification under evolving constraints. We propos

safetyarxiv-cs-ro

All Blog X Post Paper YouTube Reddit GitHub

Clear filters

Safety

Politicians who boost AI do not represent the views of their constituents. New polling in the UK shows: - Most people think AI will destroy …

DGX agent

Politicians who boost AI do not represent the views of their constituents. New polling in the UK shows: - Most people think AI will destroy more jobs than it creates (57% v 17%) - 7 in 10 are worried

safetygary-marcus--x

19 May 2026

Applications

PopPy: Opportunistically Exploiting Parallelism in Python Compound AI Applications

arXiv:2605.18697v1 Announce Type: cross Abstract: Compound AI applications, which compose calls to ML models using a general-purpose programming language like Python, are widely used for a variety of

applicationsarxiv-cs-ai

PopuLoRA: Co-Evolving LLM Populations for Reasoning Self-Play

arXiv:2605.16727v1 Announce Type: new Abstract: We introduce PopuLoRA, a population-based asymmetric self-play framework for reinforcement learning with verifiable rewards (RLVR) post-training of LLMs

agentsarxiv-cs-ai

Position: A Three-Layer Probabilistic Assume-Guarantee Architecture Is Structurally Required for Safe LLM Agent Deployment

arXiv:2605.18672v1 Announce Type: new Abstract: This position paper argues that enforcing LLM agent safety within a single abstraction layer is not merely suboptimal but categorically insufficient for

safetyarxiv-cs-ai

Position: Age Estimation Models Do Not Process Biometric Data

arXiv:2605.17347v1 Announce Type: cross Abstract: When a neural network estimates someone's age from a photograph, does it process biometric data? The answer depends on whether identity-discriminative

safetyarxiv-cs-cv

Position: AI Evaluations Should be Grounded on a Theory of Capability

arXiv:2509.19590v2 Announce Type: replace Abstract: Evaluations of generative models are now ubiquitous, and their outcomes critically shape public and scientific expectations of AI's capabilities. Ye

Position: Universal Time Series Foundation Models Rest on a Category Error

arXiv:2602.05287v2 Announce Type: replace Abstract: This position paper argues that the pursuit of 'Universal Foundation Models for Time Series' rests on a fundamental category error, mistaking a stru

agentsarxiv-cs-ai

Position: Weight Space Should Be a First-Class Generative AI Modality

arXiv:2605.18632v1 Announce Type: cross Abstract: Neural network checkpoints have quietly become a large-scale data resource: millions of trained weight vectors now exist, each encoding task-, domain-

researcharxiv-cs-ai

POST: Prior-Observation Adversarial Learning of Spatio-Temporal Associations for Multivariate Time Series Anomaly Detection

arXiv:2605.18128v1 Announce Type: new Abstract: Existing Multivariate Time Series Anomaly Detection (MTSAD) frameworks increasingly rely on integrating Graph Neural Networks (GNNs) with sequence model

Post-Trained MoE Can Skip Half Experts via Self-Distillation

arXiv:2605.18643v1 Announce Type: cross Abstract: Mixture-of-Experts (MoE) scales language models efficiently through sparse expert activation, and its dynamic variant further reduces computation by a

PPAI: Enabling Personalized LLM Agent Interoperability for Collaborative Edge Intelligence

arXiv:2605.18067v1 Announce Type: new Abstract: Deploying large language model (LLM) on edge device enables personalized LLM agents for various users. The growing availability of diverse personalized

local-aiarxiv-cs-cl

PQR: A Framework to Generate Diverse and Realistic User Queries that Elicit QA Agent Failures

arXiv:2605.16551v1 Announce Type: new Abstract: Evaluating LLM-based agents remains challenging because identifying meaningful failure cases often requires substantial human effort to design realistic

safetyarxiv-cs-cl

Preconditioned Norms: A Unified Framework for Steepest Descent, Quasi-Newton and Adaptive Methods

arXiv:2510.10777v3 Announce Type: replace Abstract: Optimization lies at the core of modern deep learning, yet existing methods often face a fundamental trade-off between adapting to problem geometry

researcharxiv-cs-lg

Predictable Confabulations: Factual Recall by LLMs Scales with Model Size and Topic Frequency

arXiv:2605.18732v1 Announce Type: cross Abstract: While scaling laws govern aggregate large language model performance, no scaling law has linked factual recall to both model size and training-data co

Applications

Prediction-Intervention Games and Invariant Sets

arXiv:2605.16828v1 Announce Type: cross Abstract: We consider the following two-player game: using observational data, the leader chooses a prediction function for a response variable Y from given cov

applicationsarxiv-cs-ai

Prediction of Challenging Behaviors Associated with Profound Autism in a Classroom Setting Using Wearable Sensors

arXiv:2605.17618v1 Announce Type: new Abstract: Autism Spectrum Disorder (ASD) is characterized by challenges with social interaction and communication and by restricted or repetitive patterns of thou

safetyarxiv-cs-ai

Predictive Prefetching for Retrieval-Augmented Generation

arXiv:2605.17989v1 Announce Type: cross Abstract: Retrieval-Augmented Generation (RAG) improves factual grounding in large language models but suffers from substantial latency due to synchronous retri

researcharxiv-cs-ai

Preference Instability in Reward Models: Detection and Mitigation via Sparse Autoencoders

arXiv:2605.16339v1 Announce Type: new Abstract: Preference learning in large language models relies on reward models as proxies for human judgment. However, these models frequently exhibit preference

researcharxiv-cs-lg

Prefix-Adaptive Block Diffusion for Efficient Document Recognition

arXiv:2605.16861v1 Announce Type: cross Abstract: Block Diffusion Models (BDMs) support parallel generation, flexible-length output, and KV caching, making them promising for efficient document parsin

researcharxiv-cs-ai

Presupposition and Reasoning in Conditionals: A Theory-Based Study of Humans and LLMs

arXiv:2605.18352v1 Announce Type: new Abstract: Presupposition projection in conditionals is central to theories of meaning and pragmatics, yet it remains largely unevaluated in large language models.

safetyarxiv-cs-cl

PriHA: A RAG-Enhanced LLM Framework for Primary Healthcare Assistant in Hong Kong

arXiv:2604.14215v2 Announce Type: replace-cross Abstract: To address the unsustainable rise in public health expenditures, the Hong Kong SAR Government is shifting its strategic focus to primary healt

PRIME: Physically-consistent Robotic Inertial and Motion Estimation for Legged and Humanoid Robots

model-releasesarxiv-cs-ro

arXiv:2605.17681v1 Announce Type: new Abstract: Humanoid and legged robots interact with the environment through intermittent contacts, making accurate motion estimation fundamentally dependent on rea

Principal Component Analysis for Lunar Crater Detection

arXiv:2605.17125v1 Announce Type: new Abstract: Optical navigation is a critical component for lunar orbiter and lander missions. Image-based crater identification has emerged as a promising technolog

researcharxiv-cs-cv

Principles of frugal inference and control

arXiv:2406.14427v4 Announce Type: replace Abstract: A central challenge for intelligent agents in an uncertain world is striking the right balance between utility maximization and resource use, not on

local-aiarxiv-cs-ai

Prior Knowledge Makes It Possible: From Sublinear Graph Algorithms to LLM Test-Time Methods

arXiv:2510.16609v3 Announce Type: replace-cross Abstract: Test-time augmentation, such as Retrieval-Augmented Generation (RAG) or tool use, critically depends on an interplay between a model's paramet

agentsarxiv-cs-ai

PRISMat: Policy-Driven, Permutation-Invariant Autoregressive Material Generation

arXiv:2605.16612v1 Announce Type: new Abstract: Rapid identification of candidate materials with target properties has become a key task in materials science. Machine learning has emerged as an altern

Privacy Policy Enforcement Guardrails for Data-Sensitive Retrieval-Augmented Generation

arXiv:2605.17034v1 Announce Type: cross Abstract: Standard PII filters often miss contextual data leakage in RAG systems, such as non-regulated attribute clusters that collectively identify individual

Privacy-Preserving Generation Fraud Detection for Distributed Photovoltaic Systems: A Solar Irradiance-Fused Federated Learning Framework

arXiv:2605.17039v1 Announce Type: new Abstract: The wide adoption of residential photovoltaic (PV) systems introduces new challenges for generation fraud detection (FD). Unlike traditional electricity

local-aiarxiv-cs-lg

Privacy Preserving Reinforcement Learning with One-Sided Feedback

arXiv:2605.18246v1 Announce Type: cross Abstract: We study reinforcement learning (RL) in multi-dimensional continuous state and action spaces with one-sided feedback, where the agent receives partial

Probing for Representation Manifolds in Superposition

arXiv:2605.18537v1 Announce Type: cross Abstract: This paper introduces the Manifold Probe, a supervised method for discovering representation manifolds in superposition. The method generalizes linear

Probing SMEFT Operators through tar{t}tar{t} Production with Hyper-Graph Neural Networks at the LHC

arXiv:2605.18382v1 Announce Type: cross Abstract: We present a phenomenological study of tar{t}tar{t} production in proton-proton collisions at sqrt{s} = 13~TeV, using a Hyper-Graph Neural Network (H-

ProfBench: Multi-Domain Rubrics requiring Professional Knowledge to Answer and Judge

arXiv:2510.18941v2 Announce Type: replace-cross Abstract: Evaluating progress in large language models (LLMs) is often constrained by the challenge of verifying responses, limiting assessments to task

Hardware

Progressive Generalization Augmentation with Deeply Coupled RND-PPO and Domain-Prioritized Noise Injection for Robust Crop Management Reinforcement Learning

arXiv:2605.17428v1 Announce Type: cross Abstract: Our preliminary experiments on gym-DSSAT maize irrigation tasks revealed that +/-2 degrees C temperature noise causes an 11.9% reduction in economic r

hardwarearxiv-cs-ai

Prompt Compression in Diffusion Large Language Models: Evaluating LLMLingua-2 on LLaDA

arXiv:2605.17932v1 Announce Type: cross Abstract: Prompt compression reduces inference cost and context length in large language models, but prior evaluations focus primarily on autoregressive archite

Prompt2Fingerprint: Plug-and-Play LLM Fingerprinting via Text-to-Weight Generation

arXiv:2605.18474v1 Announce Type: cross Abstract: The widespread deployment and redistribution of large language models (LLMs) have made model provenance tracking a critical challenge. While existing

Tutorials

PromptDecipher: Supporting AI Tutor Authoring Through Editable Simulated Interactions

arXiv:2605.16605v1 Announce Type: cross Abstract: Chatbots have long been explored as tools to support learning, and recent advances in large language models have significantly expanded the availabili

tutorialsarxiv-cs-ai

Prompts Don't Protect: Architectural Enforcement via MCP Proxy for LLM Tool Access Control

arXiv:2605.18414v1 Announce Type: cross Abstract: Large language models increasingly operate as autonomous agents that select and invoke tools from large registries. We identify a critical gap: when u

Proof-Carrying Certificates for LLM Pipelines: A Trust-Boundary Architecture

arXiv:2605.16407v1 Announce Type: cross Abstract: We present a framework for verifying the deterministic structured computations surrounding a large language model rather than the model itself, extend

agentsarxiv-cs-cl

Propagation of Chaos in Contextual Flow Maps

arXiv:2605.16747v1 Announce Type: new Abstract: We develop a quantitative statistical theory of transformers in the large-context regime by adopting the abstraction of contextual flow maps (CFMs): dyn

researcharxiv-cs-lg

PropGuard: Safeguarding LLM-MAS via Propagation-Aware Exploration and Remediation

arXiv:2605.16346v1 Announce Type: cross Abstract: LLM-based multi-agent systems (LLM-MAS) have become a promising paradigm for solving complex tasks through role specialization, tool use, memory, and

local-aiarxiv-cs-ai

PROTEA: Offline Evaluation and Iterative Refinement for Multi-Agent LLM Workflows

arXiv:2605.18032v1 Announce Type: cross Abstract: Multi-agent LLM workflows -- systems composed of multiple role-specific LLM calls -- often outperform single-prompt baselines, but they remain difficu

local-aiarxiv-cs-ai

Protection Is (Nearly) All You Need: Structural Protection Dominates Scoring in Globally Capped KV Eviction

model-releasesarxiv-cs-cl

arXiv:2605.18053v1 Announce Type: cross Abstract: We study KV cache eviction under a shared globally capped decode-time harness. Seven policies (LRU, H2O, SnapKV, StreamingLLM, Ada-KV, QUEST, Random)

Protein Fold Classification at Scale: Benchmarking and Pretraining

model-releasesarxiv-cs-lg

arXiv:2605.18552v1 Announce Type: new Abstract: Classifying protein topology is essential for deciphering biological function, but progress is held back by the lack of large-scale benchmarks that avoi

ProtoFlow: Mitigating Forgetting in Class-Incremental Remote Sensing Segmentation via Low-Curvature Prototype Flow

arXiv:2604.03212v2 Announce Type: replace Abstract: Remote sensing segmentation in real deployment is inherently continual: new semantic categories emerge, and acquisition conditions shift across seas

researcharxiv-cs-cv

ProtoSiTex: Learning Semi-Interpretable Prototypes for Multi-label Text Classification

arXiv:2510.12534v4 Announce Type: replace Abstract: The rapid growth of user-generated text across digital platforms has intensified the need for interpretable models capable of fine-grained text clas

Provable Knowledge Acquisition and Extraction in One-Layer Transformers

model-releasesarxiv-cs-cl

arXiv:2508.00901v4 Announce Type: replace-cross Abstract: Large language models may encounter factual knowledge during pre-training yet fail to reliably use that knowledge after fine-tuning. Despite g

Provably Shorter Scratchpads in Hybrid DeltaNet-Attention Decoders

model-releasesarxiv-cs-lg

arXiv:2605.16640v1 Announce Type: new Abstract: We investigate the expressive power of hybrid recurrent-attention decoders, a class of architectures used in recent open-source language models such as