safety

Phase Transitions in Driven Informational Systems: A Two-Field Perspective on Learning Theory and Non-Equilibrium Chemistry

arXiv:2605.16325v1 Announce Type: cross Abstract: Phase-transition phenomena in deep learning (grokking, emergent capabilities, and ontological reorganization under context shift) have been studied th

All Blog X Post Paper YouTube Reddit GitHub

Clear filters

Safety

Physics-Aligned Canonical Equivariant Fourier Neural Operator under Symmetry-Induced Shifts

DGX agent

arXiv:2605.18606v1 Announce Type: new Abstract: Neural operators approximate PDE solution maps, but they need not respect the symmetries of the governing equation. In out-of-distribution (OOD) regimes

safetyarxiv-cs-lg

19 May 2026

PhysioSeq2Seq: A Hybrid Physiological Digital Twin and Sequence-to-Sequence LSTM for Long-Horizon Glucose Forecasting in Type 1 Diabetes

arXiv:2605.16860v1 Announce Type: cross Abstract: Accurate long-horizon glucose forecasting is critical for automated insulin delivery systems, which help people with type 1 diabetes (T1D) manage thei

PIMSM: Physics-Informed Multi-Scale Mamba for Stable Neural Representations under Distribution Shift

arXiv:2605.16351v1 Announce Type: cross Abstract: Scientific foundation models are expected to reuse representations under changes in dataset, acquisition protocol, and deployment domain, yet many seq

Pointwise Generalization in Deep Neural Networks

arXiv:2605.18598v1 Announce Type: new Abstract: We address the fundamental question of why deep neural networks generalize by establishing a pointwise generalization theory for fully connected network

Policy-Grounded Dynamic Facet Suggestions for Job Search

arXiv:2605.16479v1 Announce Type: cross Abstract: Job seekers often initiate search with short, underspecified queries. At LinkedIn, over 80% of job-related queries contain three or fewer keywords, ma

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

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

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

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

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

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

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

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.

PULSE: Generative Phase Evolution for Non-Stationary Time Series Forecasting

arXiv:2605.16793v1 Announce Type: new Abstract: Time series forecasting under non-stationarity faces a fundamental tension between capturing stable representations and adapting to distribution shifts.

QQJ: Quantifying Qualitative Judgment for Scalable and Human-Aligned Evaluation of Generative AI

arXiv:2605.17382v1 Announce Type: new Abstract: The rapid progress of generative artificial intelligence has exposed fundamental limitations in existing evaluation methodologies, particularly for open

QuantFPFlow: Quantum Amplitude Estimation for Fokker--Planck Policy Optimisation in Continuous Reinforcement Learning

arXiv:2605.16429v1 Announce Type: cross Abstract: We introduce extbf{QuantFPFlow}, a reinforcement learning framework that integrates quantum amplitude estimation into the Fokker--Planck~(FP) formulat

Query-Conditioned Knowledge Alignment for Reliable Cross-System Medical Reasoning

arXiv:2605.18570v1 Announce Type: new Abstract: Cross-domain knowledge alignment is essential for integrating heterogeneous medical systems, yet existing approaches typically treat entity alignment as

Rad-VLSM: A Cross-Modal Framework with Semantics-Assisted Prompting for Medical Segmentation and Diagnosis

arXiv:2605.18130v1 Announce Type: new Abstract: Medical image segmentation is more clinically valuable when it supports diagnosis rather than merely producing lesion masks. However, diagnostically rel

Randomized Advantage Transformation (RAT): Computing Natural Policy Gradients via Direct Backpropagation

arXiv:2605.18591v1 Announce Type: cross Abstract: Natural policy gradients improve optimization by accounting for the geometry of distribution space, but their practical use is limited by the cost of

Ranking-Aware Calibration for Reliable Multimodal Reinforcement Learning

arXiv:2605.16999v1 Announce Type: new Abstract: Reinforcement learning post-training has substantially improved the reasoning accuracy of vision-language models, yet the resulting policies remain poor

RAVE: Re-Allocating Visual Attention in Large Multimodal Models

arXiv:2605.18359v1 Announce Type: new Abstract: Large multimodal models (LMMs) inherit the self-attention mechanism of pretrained language backbones, yet standard attention can exhibit suboptimal allo

Real-Time Aligned Reward Model beyond Semantics

arXiv:2601.22664v4 Announce Type: replace Abstract: Reinforcement Learning from Human Feedback (RLHF) is a pivotal technique for aligning large language models (LLMs) with human preferences, yet it is

Real2Sim via Active Perception with Behavior Trees Automatically Generated by VLMs

arXiv:2601.08454v2 Announce Type: replace Abstract: Constructing physically accurate simulation environments (Real2Sim) traditionally relies on manual system identification or rigid, exhaustive explor

Reasoning Before Diagnosis: Physician-Inspired Structured Thinking for ECG Classification

arXiv:2605.17308v1 Announce Type: new Abstract: Electrocardiogram (ECG) diagnosis in clinical practice relies on structured reasoning over multiple hierarchical aspects, including cardiac rhythm, cond

REC-RL: Referring expression counting via Gaussian and range-based reward optimization

arXiv:2605.16460v1 Announce Type: new Abstract: Referring expression counting (REC) is an intention-driven task that requires context-aware visual reasoning. While recent vision-language models incorp

Reducing Credit Assignment Variance via Counterfactual Reasoning Paths

arXiv:2605.16302v1 Announce Type: cross Abstract: Reinforcement learning for multi-step reasoning with large language models (LLMs) often relies on sparse terminal rewards, leading to poor credit assi

Reliability and Effectiveness of Autonomous AI Agents in Supply Chain Management

arXiv:2605.17036v1 Announce Type: new Abstract: This paper studies autonomous generative AI agents in multi-echelon supply chains using the MIT Beer Game. We identify four inference-time levers that s

Remembering More, Risking More: Longitudinal Safety Risks in Memory-Equipped LLM Agents

arXiv:2605.17830v1 Announce Type: new Abstract: Safety evaluations of memory-equipped LLM agents typically measure within-task safety: whether an agent completes a single scenario safely, often under

Representational Alignment with Chemical Induced Fit for Molecular Relational Learning

arXiv:2502.07027v2 Announce Type: replace-cross Abstract: Molecular Relational Learning (MRL) is widely applied in natural sciences to predict relationships between molecular pairs by extracting struc

Resolving Representation Ambiguity in Feedforward Novel View Synthesis Transformer via Semantic-Spatial Decoupling

arXiv:2605.18599v1 Announce Type: new Abstract: Transformer-based models have advanced feedforward novel view synthesis (NVS). Current architectures such as GS-LRM and LVSM mix semantic information (e

Responsible Federated LLMs via Safety Filtering and Constitutional AI

arXiv:2502.16691v2 Announce Type: replace Abstract: Recent research has increasingly focused on training large language models (LLMs) using federated learning, known as FedLLM. However, responsible AI

Rethinking Code Review in the Age of AI: A Vision for Agentic Code Review

arXiv:2605.17548v1 Announce Type: cross Abstract: Code review has evolved for decades, from informal peer checking to today's pull request (PR) workflows, yet it remains a largely manual, uneven, and

Retrieval and competition: how a protein foundation model starts a protein

arXiv:2605.16331v1 Announce Type: cross Abstract: Protein language models are increasingly used to guide experimental and clinical decisions, yet it is often unclear whether a confident prediction ref

Right Predictions, Misleading Explanations: On the Vulnerability of Vision-Language Model Explanations

arXiv:2605.16651v1 Announce Type: new Abstract: Explanation mechanisms are increasingly used to support transparency and trust in vision-language models (VLMs), particularly in settings where model de

RL4RLA: Teaching ML to Discover Randomized Linear Algebra Algorithms Through Curriculum Design and Graph-Based Search

arXiv:2605.18004v1 Announce Type: new Abstract: Randomized linear algebra (RLA) algorithms are a modern class of numerical linear algebra techniques that play an essential role in scientific computing

Running the same CNA search on base models (before instruction tuning) yields a structurally similar set of neurons, but ablating them produ…

Running the same CNA search on base models (before instruction tuning) yields a structurally similar set of neurons, but ablating them produces almost no behavioral change. We read this as evidence th

safetynous-research--x

S2Aligner: Pair-Efficient and Transferable Pre-Training for Sparse Text-Attributed Graphs

arXiv:2605.18579v1 Announce Type: new Abstract: Pre-training on text-attributed graphs (TAGs) is central to building transferable graph foundation models, where LLM-as-Aligner methods align graph and

SADP: Subgoal-Aware Diffusion Policy for Explainable Robots Learned from Foundation Model Generated Demonstrations

arXiv:2605.16871v1 Announce Type: new Abstract: Explainable robots require not only successful task execution but also the ability to expose internal decision-making process in a user-friendly manner.

SafeDiffusion-R1: Online Reward Steering for Safe Diffusion Post-Training

arXiv:2605.18719v1 Announce Type: new Abstract: Diffusion models have been widely studied for removing unsafe content learned during pre-training. Existing methods require expensive supervised data, e

Safety Geometry Collapse in Multimodal LLMs and Adaptive Drift Correction

arXiv:2605.18104v1 Announce Type: new Abstract: Multimodal large language models (MLLMs) often fail to transfer safety capabilities learned in the text modality to semantically equivalent non-text inp

SAPO: Step-Aligned Policy Optimization for Reasoning-Based Generative Recommendation

arXiv:2605.17648v1 Announce Type: new Abstract: Generative recommendation treats next-item prediction as autoregressive item-identifier generation. Specifically, items are encoded as semantic identifi

Scalable Bi-causal Optimal Transport via KL Relaxation and Policy Gradients

arXiv:2605.17271v1 Announce Type: cross Abstract: Bi-causal optimal transport (OT) is a natural framework for comparing and coupling stochastic processes under nonanticipative information constraints,

Scale Determines Whether Language Models Organize Representation Geometry for Prediction

arXiv:2605.17084v1 Announce Type: cross Abstract: In language models, what a representation encodes is determined by the geometry of its representation space: distances, not activations, carry meaning

Scenario Generation in Roundabouts with Adjustable Interaction Intensity

arXiv:2605.18026v1 Announce Type: new Abstract: Roundabouts, characterized by frequent merging and yielding interactions, remain a safety-critical corner case for the development and testing of intell

SD-Search: On-Policy Hindsight Self-Distillation for Search-Augmented Reasoning

arXiv:2605.18299v1 Announce Type: new Abstract: Search-augmented reasoning agents interleave internal reasoning with calls to an external retriever, and their performance relies on the quality of each

Self-Evolving Spatial Reasoning in Vision Language Models via Geometric Logic Consistency

arXiv:2605.18162v1 Announce Type: cross Abstract: Vision-Language Models (VLMs) have made striking progress, yet their spatial reasoning remains fragile: models that answer an original input correctly

Semantic Smoothing via Novel View Synthesis for Robust SAR Image Classification

arXiv:2605.16440v1 Announce Type: cross Abstract: Deep neural networks are vulnerable to adversarial perturbations, limiting deployment in safety-critical applications such as synthetic aperture radar

SG-CADVLM: A Context-Aware Decoding Powered Vision Language Model for Safety-Critical Scenario Generation

arXiv:2601.18442v3 Announce Type: replace Abstract: Autonomous Vehicle (AV) requires rigorous testing in safety-critical scenarios for safety validation, yet its validation is hindered by the high cos