Search: “safety”

Do As I Do, Not As I Say: On-Policy Distillation Inherits the Teacher in Vision-Language Models

arXiv:2610.20989v1 Announce Type: cross Abstract: Improving the reasoning capabilities of vision-language models (VLMs) often relies on post-training with supervision from stronger models. Two common

Human

Content type

All Blog X Post Paper YouTube Reddit GitHub

Clear filters

Safety

Does On-Policy Distillation Need to Complete Every Rollout?

DGX agent

arXiv:2610.21337v1 Announce Type: cross Abstract: On-policy distillation (OPD) effectively transfers capabilities from a teacher to a student through feedback on student-generated rollouts. However, g

safetyarxiv-cs-ai

16 Oct 2026

Each Reward to Its Relevant Tokens: Reward-Specific Token-Level Credit Assignment for Multi-Reward RL

arXiv:2610.21439v1 Announce Type: new Abstract: Language models are increasingly trained against several verifiable rewards at once, such as the correctness and the format of a tool call, or the indiv

Encoded but Exposed: Language Embeddings can Reveal More Than They Should in Recommender Systems

arXiv:2610.21713v1 Announce Type: cross Abstract: Language embeddings are increasingly used in recommenders to represent item content, but their privacy risks remain poorly understood. We study whethe

Enhancing Multimodal RLVR with Pairwise Token Contrast and Adaptive Trajectory Confidence

arXiv:2610.21671v1 Announce Type: new Abstract: Reinforcement learning with verifiable rewards (RLVR) improves multimodal reasoning, but group-relative policy optimization broadcasts a single sequence

EVA-Cap: Optimizing Audiovisual Video Captioning via Event-Centric Alignment

arXiv:2610.22186v1 Announce Type: cross Abstract: Recent advances in Omni Language Models (OLMs) provide a unified solution for audiovisual video captioning. Nevertheless, existing approaches predomin

Failure Is the Mother of Success: Recovering Supervision with Grounding Hindsight

arXiv:2610.21018v1 Announce Type: new Abstract: When every rollout for a problem fails, GRPO receives neither a successful exemplar nor a nonzero group-relative advantage, so it learns nothing from ex

Feasibility Ceilings Before Learning: Sizing and Timing Bounds for Drone-Mesh Interception

arXiv:2610.20823v1 Announce Type: cross Abstract: A mesh of interceptors defending a fixed site must spread over the predicted arrival point of an approaching object. This is a recurring task in multi

From Documents to Spans: Span-Centric Learning for LLM-based Explainable Medical Coding

arXiv:2603.15270v3 Announce Type: replace-cross Abstract: Medical coding, which assigns standardized ICD codes to clinical documents, is a fundamental yet challenging task in modern healthcare. It req

GeoState: Learning View-Decoupled Geometry States for 3D World Modeling

arXiv:2610.22343v1 Announce Type: cross Abstract: Recent 3D world models increasingly use features from geometry foundation models (GFMs) as predictive world states, benefiting from their rich geometr

GraphCompliance: Structure-Aware Regulatory Compliance through Policy-Context Graph Alignment

arXiv:2510.26309v2 Announce Type: replace Abstract: Regulatory compliance is an important requirement in many industrial and administrative settings. Yet compliance assessment remains difficult becaus

GRID-Planner: Graph-Structured Relational Reward Shaping for Diffusion Planning

arXiv:2610.21984v1 Announce Type: cross Abstract: Autonomous driving planning requires selecting trajectories that remain safe, rule-compliant, and comfortable under complex multi-agent interactions,

Honesty is the Best Policy ... or is it? Demand Management for Probabilistic Service Options

arXiv:2610.21351v1 Announce Type: new Abstract: To enable well-informed customer choices and redistribute demand across service options, providers can display probabilities regarding predicted service

How Do LLM-Generated Texts Impact Term-Based Retrieval Models?

arXiv:2508.17715v2 Announce Type: replace-cross Abstract: As more content generated by large language models (LLMs) floods into the Internet, information retrieval (IR) systems now face the challenge

Identification and Inference for Algorithmic Frontiers with Selective Labels

arXiv:2606.14977v2 Announce Type: replace-cross Abstract: This paper provides identification results to characterize a fairness-accuracy (FA) frontier, and statistical inference tools to test hypothes

KnowGPO: Knowledge-Grouped Policy Optimization for Search Agents

arXiv:2610.22331v1 Announce Type: new Abstract: Deep search agents solve knowledge-intensive tasks through long-horizon, multi-turn interactions, where assigning credit to individual search decisions

KODA: Kinematics--Observation Data Alignment for Humanoid Learning from Human Egocentric Demonstration

arXiv:2610.21152v1 Announce Type: new Abstract: Training humanoid whole-body vision-language-action (VLA) policies requires costly robot demonstrations with aligned robot-view observations and control

Learning at the Right Pace: Adaptive Data Scheduling Improves LLM Reinforcement Learning

arXiv:2606.22305v2 Announce Type: replace Abstract: Large Language Models (LLMs) achieve remarkable reasoning capabilities through reinforcement learning (RL) post-training. However, existing RL post-

Learning from Unverified Peers via Shared Residual Distillation

arXiv:2610.22080v1 Announce Type: new Abstract: A language model can generate several solutions to the same problem, some leading to correct answers and others to incorrect ones. Many post-training me

Learning Intrinsic Dynamics in Latent World Models

arXiv:2610.21612v1 Announce Type: new Abstract: Learning intrinsic dynamics---the relationship between actions and their consequences---is essential for world models. However, visual variations unrela

Learning Rewards in Strategic Games

arXiv:2610.21912v1 Announce Type: cross Abstract: Understanding the objectives that drive agents' behavior is crucial in strategic environments, where multiple agents interact and adapt to one another

Learning to Forecast on Multiple Spatiotemporal Hierarchies

arXiv:2610.21414v1 Announce Type: new Abstract: Hierarchical forecasting exploits aggregation structures that expose multiple views of the same underlying spatiotemporal forecasting problem, while ens

Learning to Reason with Selective Internal Feedback

arXiv:2610.21392v1 Announce Type: new Abstract: Reinforcement learning from internal feedback can improve language models without ground-truth answers or external verifiers. However, confidence and co

Learning to Steer LLM Agents and Multi-Agent Systems

arXiv:2610.22316v1 Announce Type: cross Abstract: Large language model (LLM)-based agents and multi-agent systems increasingly tackle complex tasks through long-horizon tool use and inter-agent collab

Learning Topology-Aware Representations for Mechanics-Relevant TPMS Metamaterials

arXiv:2610.20684v1 Announce Type: cross Abstract: Learning informative representations of mechanical Triply Periodic Minimal Surface (TPMS) metamaterials is important for mechanics analysis and struct

Learning Warm Starts for the Simplex Method from Measured Solver Cost

arXiv:2610.20892v1 Announce Type: cross Abstract: Learned warm starts for the simplex method build starting bases using predictions of which variables will be basic at optimality. Prior work has shown

Line-Augmented cuVSLAM: Robust Visual Odometry in Low-Texture Environments

arXiv:2610.20958v1 Announce Type: cross Abstract: Autonomous mobile robots in warehouses and production facilities operate in precisely the environments that defeat point-feature visual odometry: feat

MAIA: Anticipating Interaction Anomalies in Contact-Rich Manipulation via Multimodal World Models

arXiv:2610.21686v1 Announce Type: new Abstract: Reliable contact-rich manipulation requires detecting execution deviations early enough for corrective action. Yet most monitoring methods are reactive,

MARS: A Momentum-Aware Adaptive Rating System for Personalized Assessment

arXiv:2610.21685v1 Announce Type: cross Abstract: Adaptive assessment systems personalize learning by estimating learner proficiency and recommending suitable questions. However, many existing approac

Matoi: Selective Revalidation of Consequence Evidence for Agent Authorization

arXiv:2610.21394v1 Announce Type: cross Abstract: Matoi separates reusable consequence evidence from current execution authority and durable occurrence history. Its finite query-aware contract covers

MedPAO-VL: Interpretable and Protocol-Compliant Agent for Structured Radiology Report Generation

arXiv:2610.21794v1 Announce Type: new Abstract: Despite the remarkable generalization capabilities of Large Language Models, generating clinically accurate and structured radiology reports from medica

Mem4Act: Learning Effective Latent Memory for Long-Horizon Robotic Manipulation

arXiv:2610.22323v1 Announce Type: new Abstract: Long-horizon robotic manipulation fundamentally requires memory: agents must retain task-relevant information from past observations and interactions an

MemComposer: Learning to Assemble Memories for LLM Agents via Reinforcement Learning

arXiv:2610.21091v1 Announce Type: new Abstract: Agent memory enables large language model (LLM) agents to reuse past experiences without updating their parameters. However, semantic similarity alone d

Mind the Shift: Decoding Monetary Policy Stance from FOMC Statements with Large Language Models

arXiv:2603.14313v2 Announce Type: replace Abstract: Federal Open Market Committee (FOMC) statements are a major source of monetary-policy information, and even subtle changes in their wording can move

Mine, Sample, Render: Test-Time Diversity Alignment Of Generative Models

arXiv:2610.21185v1 Announce Type: new Abstract: Modern generative models produce high-quality individual outputs but struggle to sample diversely, even when explicitly instructed to do so. We introduc

MLExpert: Towards a Generalist Agentic Model for Autonomous Machine Learning Engineering

arXiv:2610.22344v1 Announce Type: new Abstract: Large language models (LLMs) are expanding from assisting scientific research and engineering to autonomously developing and improving AI systems, advan

MMSearch-OPD: Easy Made Hard, Hard Made Learnable for Long-Horizon Multimodal Search Agents

arXiv:2610.22240v1 Announce Type: new Abstract: Multimodal deep search is moving from short queries to long-horizon investigations that continually acquire evidence. However, training such agents rema

MoCaP: Mobile Code-as-Policies for Zero-Shot Mobile Manipulation from Passive Human Videos

arXiv:2610.20998v1 Announce Type: new Abstract: Passive human videos of everyday activity offer robots access to a massive amount of human experience without requiring a demonstration for each task. H

Monotone Piecewise-Linear Harmonization of Multitemporal Multispectral Surface Reflectance With Unsupervised Deep Multiview Pseudoinvariant Features Learning

arXiv:2610.21183v1 Announce Type: cross Abstract: Surface reflectance products derived from optical satellite may exhibit systematic inconsistencies across acquisition dates, because atmospheric corre

MSP: Modality Self-Play for Training Multimodal LLMs Without Paired Cross-Modal Data

arXiv:2610.20965v1 Announce Type: new Abstract: Large multimodal models have achieved remarkable progress by scaling model capacity and training on massive paired multimodal data. However, this paradi

Multi-Teacher On-Policy Distillation for Flow-Based Vision-Language-Action Models

arXiv:2610.21522v1 Announce Type: new Abstract: Specialized vision-language-action (VLA) policies acquire strong domain-specific skills, yet consolidating these skills into one shared policy remains c

Neither True nor False: Do LLMs Represent Truth-Value Gaps in Plural Predication?

arXiv:2610.14438v2 Announce Type: replace Abstract: Large Language Models (LLMs) are increasingly being used to summarize and verify claims in critical sources that inform consequential decisions. Suc

NeoHorse-1: Towards Recursive Self-Improvement via Agentic Post-Training with Routing Harness

arXiv:2609.08183v2 Announce Type: replace Abstract: Recursive self-improvement (RSI) requires a concrete mechanism through which an AI system observes its capabilities and converts that evidence into

Noise-Space Reward Alignment for Few-Step Generative Models

arXiv:2610.21125v1 Announce Type: new Abstract: Few-step generative models produce high-quality samples in one or a few evaluations. However, steering them toward desired properties remains difficult,

not the first time Dwarkesh comes as off to me as working as an undisclosed press secretary for the tech industry.

not the first time Dwarkesh comes as off to me as working as an undisclosed press secretary for the tech industry. Dwarkesh Patel reveals an Anthropic employee told him two days ago they expect AIs to

safetygary-marcus--x

OmniVerifier-C1: Decomposed Graded Verifier for Rule-Governed Generative World Modeling

arXiv:2610.22355v1 Announce Type: new Abstract: Generative visual verifiers play a crucial role in advancing multimodal intelligence. To enhance the robustness of visual verifiers in complex multimoda

On-Policy Residual Learning for Diffusion Language Models

arXiv:2610.20687v1 Announce Type: new Abstract: Residual context is key to unlocking the speed-quality Pareto frontier of diffusion language models (DLMs) by carrying forward predictive distributions

One Bias Score, Two Factors: How Mitigations Affect Option Bias in Multiple-Choice Evaluation

arXiv:2610.20993v1 Announce Type: new Abstract: Reordering multiple-choice options can change a model's answer, an effect studied as selection bias (a preference for certain option letters) and as ord