Search: “safety”

How Off-Policy Can GRPO Be? Mu-GRPO for Efficient LLM Reinforcement Learning

arXiv:2605.17570v1 Announce Type: cross Abstract: Group Relative Policy Optimization (GRPO) has been a key driver of recent progress in reinforcement learning with verifiable rewards (RLVR) for large

All Blog X Post Paper YouTube Reddit GitHub

Clear filters

How to Instruct Your Robot: Dense Language Annotations Power Robot Policy Learning

DGX agent

arXiv:2605.17077v1 Announce Type: cross Abstract: Scaling robot policy learning is bottlenecked by the cost of collecting demonstrations, while language annotations for existing demonstrations are com

safetyarxiv-cs-ai

19 May 2026

How Wrong Can Your Counterfactual Be? Quantifying Confounding Bias for Continuous Treatments without a Control Group

arXiv:2603.07438v2 Announce Type: replace Abstract: Stress testing poses a causal question: how would portfolio credit losses change if the macroeconomy followed an adverse counterfactual path? Yet st

Identifying Latent Actions and Dynamics from Offline Data via Demonstrator Diversity

arXiv:2603.17577v2 Announce Type: replace-cross Abstract: Can latent actions and environment dynamics be recovered from offline trajectories when actions are never observed? We study this question in

'I'm Not Mad, Just Focused'': Understanding Human Emotions in Human-Robot Collaboration

arXiv:2605.16816v1 Announce Type: new Abstract: Human-robot collaboration (HRC) can benefit from robots' abilities to interpret human emotional states. However, current emotion recognition (ER) models

safetyarxiv-cs-ro

Implicit Hierarchical GRPO: Decoupling Tool Invocation from Execution for Tool-Integrated Mathematical Reasoning

arXiv:2605.18500v1 Announce Type: new Abstract: Large language models (LLMs) have increasingly leveraged tool invocation to enhance their reasoning capabilities. However, existing approaches typically

Improved Baselines with Representation Autoencoders

arXiv:2605.18324v1 Announce Type: cross Abstract: Representation Autoencoders (RAE) replace traditional VAE with pretrained vision encoders. In this paper, we systematically investigate several design

Improving MLLM Training Efficiency via Stage-Aware Sparsity

arXiv:2509.18150v2 Announce Type: replace-cross Abstract: Multimodal Large Language Models (MLLMs) have demonstrated outstanding performance across a variety of domains. However, training MLLMs is oft

Individual utilities of life satisfaction reveal inequality aversion unrelated to political alignment

arXiv:2509.07793v4 Announce Type: replace-cross Abstract: How should well-being be prioritised in society, and what trade-offs are people willing to make between fairness and personal well-being? We i

InFeR: Informed Failure Resilience in Learned Visual Navigation Control

arXiv:2510.24680v2 Announce Type: replace Abstract: While imitation learning (IL) has enabled successful visual navigation in many common environments, IL policies are prone to unpredictable failures

safetyarxiv-cs-ro

Interpretable epistemic uncertainty decomposition in sequential generative models via polynomial chaos surrogates

arXiv:2510.21523v2 Announce Type: replace Abstract: Sequential generative models conditioned on uncertain rewards are central to AI-driven scientific discovery, yet the epistemic uncertainty they inhe

it’s strange how many podcasters shy away from airing both sides of an argument that may totally shape our lives. https://x.com/benjamin_hor…

Gary Marcus critiques podcasters for avoiding balanced discussion of contentious issues that significantly impact society, suggesting there is an reluctance to present opposing viewpoints on important

safetygary-marcus--x

LACE: Latent Visual Representation for Cross-Embodiment Learning

arXiv:2605.16743v1 Announce Type: new Abstract: Cross-embodiment learning from human demonstrations is hindered by the visual gap between human and robot embodiments. While self-supervised learning (S

safetyarxiv-cs-ro

Lance: Unified Multimodal Modeling by Multi-Task Synergy

arXiv:2605.18678v1 Announce Type: cross Abstract: We present Lance, a lightweight native unified model supporting multimodal understanding, generation, and editing for both images and videos. Rather t

Latent Action Control for Reasoning-Guided Unified Image Generation

arXiv:2605.16961v1 Announce Type: cross Abstract: Unified multimodal models can encode visual understanding and image generation within a shared backbone, yet understanding does not automatically tran

LatentUMM: Dual Latent Alignment for Unified Multimodal Models

arXiv:2605.17766v1 Announce Type: new Abstract: Unified multimodal models (UMMs) achieve strong performance in both understanding and generation by learning a shared latent space, yet they often exhib

Learning Fill-in Reduction Ordering via Graph Policy Optimization for Sparse Matrices

arXiv:2605.17362v1 Announce Type: new Abstract: Matrix reordering in large sparse solvers seeks a permutation that minimizes factorization fill-in to reduce memory and computation. Because the minimum

Learning Multi-Timescale Abstractions for Hierarchical Combinatorial Planning

arXiv:2605.17058v1 Announce Type: new Abstract: The combination of exponentially large action spaces, stochastic dynamics, and long-horizon decision-making under limited resources makes Sequential Sto

Learning Native Continuation for Action Chunking Flow Policies

arXiv:2602.12978v2 Announce Type: replace-cross Abstract: Action chunking enables Vision Language Action (VLA) models to run in real time, but naive chunked execution often exhibits discontinuities at

Learning Relative Representations for Fine-Grained Multimodal Alignment with Limited Data

arXiv:2605.16834v1 Announce Type: cross Abstract: Multimodal pre-training demonstrates strong generalization performance, but this paradigm is often impractical in domains where paired data are scarce

Learning to Reason without External Rewards

arXiv:2505.19590v5 Announce Type: replace-cross Abstract: Training large language models (LLMs) for complex reasoning via Reinforcement Learning with Verifiable Rewards (RLVR) is effective but limited

Learning Transferable Topology Priors for Multi-Agent LLM Collaboration Across Domains

arXiv:2605.17359v1 Announce Type: new Abstract: Large language model (LLM)-based multi-agent systems have shown strong potential for complex reasoning by coordinating specialized agents through struct

Learning under Distributional Drift: Prequential Reproducibility as an Intrinsic Statistical Resource

arXiv:2512.13506v4 Announce Type: replace Abstract: Statistical learning under distributional drift remains poorly characterized, especially in closed-loop settings where learning alters the data-gene

LegalCheck: Retrieval- and Context-Augmented Generation for Drafting Municipal Legal Advice Letters

arXiv:2605.12012v2 Announce Type: replace Abstract: Public-sector legal departments in the Netherlands face acute staff shortages, increased case volumes, and increased pressure to meet regulatory com

Linguistic Uncertainty and Reply Engagement on X: A Cross-Domain Replication of the Uncertainty-Reply Asymmetry

arXiv:2605.16289v1 Announce Type: cross Abstract: Linguistic uncertainty is common in social media, but its relationship with engagement remains unclear across languages and topics. Using 2,258 Englis

LISTEN to Your Preferences: An LLM Framework for Multi-Objective Selection

arXiv:2510.25799v2 Announce Type: replace Abstract: Human experts often struggle to select the best option from a large set of items with multiple competing objectives, a process bottlenecked by the d

Mamba-VGGT: Persistent Long-Sequence Video Geometry Grounded Transformer via External Sliding Window Mamba Memory

arXiv:2605.17478v1 Announce Type: new Abstract: Visual Geometry Grounded Transformers (VGGT) have set new benchmarks in high-fidelity 3D scene reconstruction. However, as the sequence length increases

MARR: Module-Adaptive Residual Reconstruction for Low-Bit Post-Training Quantization

arXiv:2605.17997v1 Announce Type: cross Abstract: Recently, residual reconstruction-based model quantization methods have achieved promising performance in low-bit post-training quantization (PTQ) by

Masking Causality and Conditional Dependence

arXiv:2603.06984v2 Announce Type: replace-cross Abstract: Many regulatory and analytic problems require that a prohibited variable influence a decision only through a designated allowable channel -- a

Measuring Changes in Instructor Class Design and Student Learning After the Release of Large Language Models (LLMs)

arXiv:2605.16284v1 Announce Type: cross Abstract: Student use of Generative AI (GenAI) products in completing their classwork, with or without their professors' knowledge and/or approval, has resulted

Medical Context Distorts Decisions in Clinical Vision Language Models

arXiv:2605.17436v1 Announce Type: cross Abstract: Vision-language models (VLMs) are increasingly proposed for clinical decision support, yet their reliability in real-world scenarios that require inte

Minor First, Major Last: A Depth-Induced Implicit Bias of Sharpness-Aware Minimization

arXiv:2603.08290v2 Announce Type: replace-cross Abstract: We study the implicit bias of Sharpness-Aware Minimization (SAM) when training L-layer linear diagonal networks on linearly separable binary c

Mitigating Conversational Inertia in Multi-Turn Agents

arXiv:2602.03664v3 Announce Type: replace Abstract: Large language models excel as few-shot learners when provided with appropriate demonstrations, yet this strength becomes problematic in multiturn a

MoASE++: Mixture of Activation Sparsity Experts with Domain-Adaptive On-policy Distillation for Continual Test Time Adaptation

arXiv:2605.17743v1 Announce Type: new Abstract: Continual test-time adaptation adapts a source-pretrained model to non-stationary, unlabeled target streams while retaining past competence, yet texture

MSIQ: Moment-based Scale-Invariant Quality Measure for Single Image Super-Resolution

arXiv:2605.17588v1 Announce Type: new Abstract: Assessing the quality of single image super-resolution (SISR) results remains an open methodological problem. Common full-reference metrics (PSNR, SSIM,

Multi-Order Matching Network for Alignment-Free Depth Super-Resolution

arXiv:2511.16361v3 Announce Type: replace Abstract: Recent guided depth super-resolution methods are premised on the assumption of strict spatial alignment between depth and RGB, achieving high-qualit

Multilingual and Multimodal LLMs in the Wild: Building for Low-Resource Languages

arXiv:2605.17152v1 Announce Type: new Abstract: Multimodal LLMs are evolving from vision-language to tri-modality that see, hear, and read, yet pipelines and benchmarks remain English-centric and comp

Model Releases

Multilingual jailbreaking of LLMs using low-resource languages

model-releasesarxiv-cs-ai

arXiv:2605.18239v1 Announce Type: cross Abstract: Large Language Models (LLMs) remain vulnerable to jailbreak attempts that circumvent safety guardrails. We investigate whether multi-turn conversation

Natural-Language Agent Harnesses

arXiv:2603.25723v2 Announce Type: replace-cross Abstract: Agent performance is strongly shaped by the surrounding harness: the external execution system around a model that organizes a task run. Yet t

Neural equilibria for long-term prediction of nonlinear conservation laws

arXiv:2501.06933v3 Announce Type: replace Abstract: Nonlinear conservation laws govern a broad class of important physical systems in science and industry and are central to scientific machine learnin

Neural Visual Decoding via Cognitive guided Adaptive Blurring and Information Constrained Alignment

arXiv:2605.16418v1 Announce Type: cross Abstract: EEG-based visual decoding aims to establish a mapping between neural signals and visual semantics. However, it remains constrained by the dual challen

NEWTON: Agentic Planning for Physically Grounded Video Generation

arXiv:2605.18396v1 Announce Type: new Abstract: Video generation models produce visually compelling results but systematically violate physical commonsense -- on VideoPhy-2, the best model achieves on

Old Habits Die Hard: How Conversational History Geometrically Traps LLMs

arXiv:2603.03308v2 Announce Type: replace-cross Abstract: How does the conversational past of large language models (LLMs) influence their future performance? Recent work suggests that LLMs are affect

oldsymbol{f}-OPD: Stabilizing Long-Horizon On-Policy Distillation with Freshness-Aware Control

arXiv:2605.17862v1 Announce Type: cross Abstract: Scaling on-policy distillation (OPD) for large language models (LLMs) confronts a fundamental tension: asynchronous execution is necessary for system

On the Adversarial Robustness of Large Vision-Language Models under Visual Token Compression

arXiv:2601.21531v2 Announce Type: replace-cross Abstract: Visual token compression is widely used to accelerate large vision-language models (LVLMs) by pruning or merging visual tokens, yet its advers

OPERA: A Reinforcement Learning--Enhanced Orchestrated Planner-Executor Architecture for Reasoning-Oriented Multi-Hop Retrieval

arXiv:2508.16438v4 Announce Type: replace-cross Abstract: Recent advances in large language models (LLMs) and dense retrievers have driven significant progress in retrieval-augmented generation (RAG).

Optimal Control of Multiclass Fluid Queueing Networks: A Machine Learning Approach

arXiv:2307.12405v2 Announce Type: replace Abstract: We propose a machine learning approach to the optimal control of multiclass fluid queueing networks (MFQNETs) that provides explicit and insightful

OrbiSim: World Models as Differentiable Physics Engines for Embodied Intelligence

arXiv:2605.16395v1 Announce Type: cross Abstract: We present OrbiSim, a novel robotic simulation paradigm that redefines world models as a fully differentiable physics engine for embodied intelligence