Search: “safety”

Learning Perceptual Geometry of Natural Odor Mixtures from LLM-Derived Descriptors

arXiv:2610.17164v1 Announce Type: new Abstract: The relationship between odor chemistry and perception remains incompletely understood. Many existing studies focus on the perception of single molecule

Content type

All Blog X Post Paper YouTube Reddit GitHub

Clear filters

Lifted State Policy Optimization: Internalizing Self-Correction through Transition Geometry

DGX agent

arXiv:2610.20350v1 Announce Type: cross Abstract: Preference optimization ranks final answers and trajectory RL maximizes aggregate returns, yet neither credits the local change that turns a flawed re

safetyarxiv-cs-lg

15 Oct 2026

Long-Horizon RL Needs Muon

arXiv:2610.20256v1 Announce Type: new Abstract: Long-horizon agents must coordinate reasoning, tool use, and environment feedback over extended interactions. Learning from sparse outcome rewards requi

Lost in Translation: Why Lower Bias Scores Don't Mean Fairer Models

arXiv:2610.20479v1 Announce Type: new Abstract: Bias audits of multilingual language models usually compare a model's average stereotype score in English with its average in another language and read

MAEM: Mitigating Recency Bias via a Replay-Free, Capacity-Controlled Dynamic Mixture of Experts for Class-Incremental Graph Learning

arXiv:2610.20238v1 Announce Type: new Abstract: Graph Neural Networks (GNNs) excel on graph-structured data but suffer catastrophic forgetting when classes arrive sequentially. In class-incremental gr

Maximum-Entropy Exploration with Future State-Action Visitation Measures

arXiv:2603.18965v2 Announce Type: replace Abstract: Maximum entropy reinforcement learning motivates agents to explore states and actions to maximize the entropy of some distribution, typically by pro

MemDuet: Learning Dual-Process Memory for Long-Context Reasoning

arXiv:2610.19928v1 Announce Type: new Abstract: Reasoning over infinitely long documents is becoming a reality with recurrent memory agents. However, increasing context length exacerbates the trade-of

MIPS: Mitigating Mode Collapse in Reinforcement Learning for Large Outcome Spaces

arXiv:2610.19457v1 Announce Type: new Abstract: Expected-reward maximization in reinforcement learning concentrates probability mass on a few outcomes, even in tasks that require many distinct, high-q

MMLA-RSI: Recursive Self-Improvement through RTT and RTMA

arXiv:2610.20294v1 Announce Type: new Abstract: An agent can improve its behavior within an unresolved problem by changing how it acts and which experience it retains. We propose MMLA-RSI, a recursive

Model What Matters: Decision-Guided Selective World Modeling for Language Agents

arXiv:2610.19642v1 Announce Type: cross Abstract: Reinforcement learning provides a direct mechanism for optimizing task objectives in large language model-driven agents, but task rewards are typicall

Multi-Sensitive Distributional Fairness: Joint Wasserstein Audits and Shapley Allocation

arXiv:2610.19252v1 Announce Type: cross Abstract: Fairness audits with several sensitive attributes are often performed one coordinate at a time. This can miss purely intersectional disparities: a sco

Neos: Learning to Interleave Search and Visual Reasoning

arXiv:2610.17043v1 Announce Type: new Abstract: Recent multimodal deep search agents have advanced tool use and visual content retrieval, yet they still struggle to convert intermediate visual evidenc

NePPO: Near-Potential Policy Optimization for General-Sum Multi-Agent Reinforcement Learning

arXiv:2603.06977v3 Announce Type: replace Abstract: Multi-agent reinforcement learning (MARL) is increasingly used to design learning-enabled agents that interact in shared environments. However, trai

Neural Supervised Clustering for Learning Individualized Treatment Rules with Many Treatments

arXiv:2610.19104v1 Announce Type: cross Abstract: Learning an individualized treatment rule (ITR) with many treatments requires estimating treatment-specific effects from limited observations per arm.

Noisy-Label Calibration with Multiple Unreliable Labels

arXiv:2610.19543v1 Announce Type: cross Abstract: In medical annotation and crowdsourcing, multiple annotators assign labels to each example, but these labels may be incorrect. Treating such noisy lab

Not All Criteria Matter: Importance-Aware Criteria Weighting for Rubric-Based LLM Evaluation

arXiv:2610.19915v1 Announce Type: new Abstract: Rubric-based evaluation assesses the responses of large language models (LLMs) by aggregating the verification results of explicit criteria, yet not all

Occupancy Modeling via Drifting: One-Step Generation Enables Direct Policy Extraction

arXiv:2610.19883v1 Announce Type: new Abstract: The growing interest in behavior foundation models, combined with recent advancements in generative modeling, has led to an appealing recipe for reinfor

Of Errors and Echoes: Responsibility-Aware Credit Assignment for Tool-Use Agents

arXiv:2610.16027v1 Announce Type: cross Abstract: Reinforcement learning improves tool-use agents, yet fine-grained credit assignment remains challenging even with step-level feedback. An observed err

On the Complexity Gap Between POMDPs and Multi-Environment POMDPs

arXiv:2610.15010v1 Announce Type: new Abstract: Multi-environment POMDPs (ME-POMDPs) generalize POMDPs by encoding a finite set of possible environments, capturing model uncertainty. The finite-horizo

On the Convergence of Semi-Gradient TD: A Mean Field Perspective

arXiv:2610.20014v1 Announce Type: new Abstract: Semi-gradient temporal-difference learning is a core algorithmic component of reinforcement learning, but bootstrapping with function approximation gene

On the Limits of Representation-Based Test-Time Adaptation for Multimodal LLMs

arXiv:2610.19892v1 Announce Type: cross Abstract: Vision-language models suffer significant accuracy drops under common corruptions. While test-time adaptation (TTA) successfully mitigates these shift

One-shot Human Video-Guided Adaptation for Vision-Language-Action Models

arXiv:2610.19746v1 Announce Type: new Abstract: Vision-language-action (VLA) models have recently achieved remarkable progress by leveraging large-scale robot datasets and foundation models for genera

One Task, Many Interfaces: Simple Cross-Interface Consistency Training Improves LLM Agent Robustness

arXiv:2610.20576v1 Announce Type: cross Abstract: Post-training binds the competence of an LLM agent to the interface it was trained on. Semantics-preserving perturbations of that interface, such as r

OneLA: Learning Many Worlds from Latent Actions

arXiv:2610.20160v1 Announce Type: new Abstract: Learning world models across diverse domains calls for a unified control space, yet ac- tion interfaces vary across domains and action annotations are o

Outcome-Routed Distillation and Reinforcement Learning for Diffusion Language Models

arXiv:2610.19271v1 Announce Type: new Abstract: On-policy distillation provides dense supervision on the student's own samples, but its targets still come from the teacher. Verifier reinforcement lear

Outperforming 1000 Euclidean Layers With Hyperbolic Goal-Conditioned RL

arXiv:2610.20009v1 Announce Type: new Abstract: Scaling self-supervised contrastive reinforcement learning (CRL) to 1000-layer Euclidean neural networks has enabled state-of-the-art goal-reaching perf

PA1-OPD: One-Step Prefix-Augmented On-Policy Distillation for Language Model Pretraining

arXiv:2610.19601v1 Announce Type: new Abstract: Knowledge distillation (KD) improves language-model pretraining by matching teacher distributions at corpus prefixes, but provides no direct supervision

Paraphrase Consensus Is Not Self-Consistency: Instruction Reliance and Robustness in VLA Policies

arXiv:2610.19438v1 Announce Type: new Abstract: Vision-language-action (VLA) policies fine-tuned on LIBERO are trained and evaluated on the same ten instruction strings, and rewording the instruction

PEARL: Process-Experience Aware Reinforcement Learning for Multi-Turn Agents

arXiv:2610.14636v1 Announce Type: new Abstract: Terminal rewards make multi-turn agents trainable without step-level annotations, but group-relative advantages provide little guidance about which inte

PhySGA: A Physics-Space Grounded Agentic Framework for Generalizable and Interactive 3D Composition

arXiv:2610.20560v1 Announce Type: cross Abstract: 3D object composition is pivotal to digital content creation and robot simulation, while also serving as a critical testbed for physical intelligence

Physical Time as a Missing Dimension of Robot Policy Memory Dynamics

arXiv:2610.20524v1 Announce Type: new Abstract: Existing robot policy memories typically evolve with observation sequences, implicitly treating each update as the same amount of physical time. Yet ide

Physics-Shaped, Not Physical: Auditing Physics-Guided Underwater Image Restoration

arXiv:2610.19763v1 Announce Type: new Abstract: Physics-guided networks for underwater image restoration predict transmission and backscatter and invert the image-formation model, yet rarely test whet

PolyMorph: Learning Locomotion Policy for Cross-Embodiment Adaptation

arXiv:2610.19615v1 Announce Type: new Abstract: Reinforcement learning has enabled robust locomotion for legged robots, but learned policies are typically tied to a specific robot morphology and requi

PopGen: Generating Semantically Diverse Populations for Zero-Shot Coordination

arXiv:2610.15969v1 Announce Type: new Abstract: In multi-agent reinforcement learning, exposing a cooperator to partners with diverse strategies during training is key to better collaboration at test

Post-Training, Finetuning, and Steering:A Field Guide for Curating Training Data

arXiv:2610.19143v1 Announce Type: new Abstract: While Vision-Language-Action models have enabled impressive cross-task transfer in zero-shot settings, adaptation methods are often necessary to achieve

Pre-Trained Language Models as Zero-Shot Tools for Social Science Research

arXiv:2610.19048v1 Announce Type: new Abstract: Record linkage and demographic inference from names are routine demands in social science. Most approaches to these problems require labeled training da

RAP-WAM: Learning World-Action Models through Predictive Rehearsal and Action Perturbations

arXiv:2610.16443v1 Announce Type: new Abstract: Recent world action models draw on video generation priors to jointly predict future observations and robot actions. Adapting these priors to control re

ReAl: Refine neural operators via aligning dynamical flow

arXiv:2610.20369v1 Announce Type: new Abstract: Neural partial differential equation (PDE) surrogates trained with mean squared error (MSE) can capture dominant dynamics but often lose high-frequency

Recovering Reliable Supervision via Critique-Guided On-Policy Distillation

arXiv:2610.20572v1 Announce Type: new Abstract: On-policy distillation (OPD) trains a student model on its own rollouts with token-level supervision from a teacher, assuming that the teacher can reach

Refine Actions If Needed: Adaptive Action Chunk Refinement for Robot Manipulation

arXiv:2610.19765v1 Announce Type: new Abstract: Computationally lightweight manipulation policies predict executable actions with low inference latency, but prediction errors can limit task success. I

Reinforcement Learning via Advantage Guidance

arXiv:2610.19783v1 Announce Type: new Abstract: Reinforcement learning (RL) holds the promise of learning a policy that can outperform what is shown in the data. However, in practice, such systems suf

Relational Semantic Grounding for Test-Time Discovery

arXiv:2610.14721v1 Announce Type: new Abstract: Test-Time Discovery (TTD) aims to recognize known classes and detect novel ones from a continuous unlabeled test stream after deployment. Existing TTD m

Rent-Seeking Language Models: Reward Models Set the Prices, Frequent Audits Curb the Rent

arXiv:2610.19832v1 Announce Type: new Abstract: Reward model overoptimization in RLHF is usually described as reward hacking. We recast it as rent-seeking: the policy, a firm, earns reward from a regu

Representation Conditioning for Generation

arXiv:2610.20608v1 Announce Type: new Abstract: Representation-alignment methods effectively accelerate the diffusion transformer training. REPA, for example, uses a pretrained vision encoder to super

RetailAgent: Structured Adverse Timing in Self-Conditioned Multimodal LLM Trading Agents

arXiv:2608.28399v2 Announce Type: replace Abstract: In financial markets, a sequential policy that reacts systematically to price movements may become predictable to other market participants. This pa

Retell, Reward, Repeat: Reinforcement Learning for Narrative Theory-Informed Story Retelling

arXiv:2601.17226v3 Announce Type: replace-cross Abstract: Counterfactual storytelling exposes LLM shortcomings in causal reasoning, where few-shot prompting and SFT fail to produce logical and rationa

RL Post-Training Diffusion Language Models Along Their Own Generation Paths

arXiv:2610.17266v1 Announce Type: new Abstract: Diffusion language models are a promising and more efficient alternative to autoregressive models. However, most RL post-training techniques are not eas

RouteWeaver: Weaving Mode Selection and Execution into Unified LLM Routing

arXiv:2610.19705v1 Announce Type: new Abstract: LLM routing dynamically selects and coordinates heterogeneous language models across diverse tasks. However, existing approaches are largely built aroun