Search: “safety”

Do Not Step Into the Same River Twice: Learning to Reason from Trial and Error

arXiv:2510.26109v4 Announce Type: replace Abstract: Reinforcement learning with verifiable rewards (RLVR) has significantly boosted the reasoning capability of language models (LMs). However, existing

All Blog X Post Paper YouTube Reddit GitHub

Clear filters

Safety

DockAnywhere: Data-Efficient Visuomotor Policy Learning for Mobile Manipulation via Novel Demonstration Generation

DGX agent

arXiv:2604.15023v1 Announce Type: new Abstract: Mobile manipulation is a fundamental capability that enables robots to interact in expansive environments such as homes and factories. Most existing app

safetyarxiv-cs-ro

17 Apr 2026

Emergent Neural Automaton Policies: Learning Symbolic Structure from Visuomotor Trajectories

arXiv:2603.25903v2 Announce Type: replace Abstract: Scaling robot learning to long-horizon tasks remains a formidable challenge. While end-to-end policies often lack the structural priors needed for e

safetyarxiv-cs-ro

Enhancing LLM-based Search Agents via Contribution Weighted Group Relative Policy Optimization

arXiv:2604.14267v1 Announce Type: new Abstract: Search agents extend Large Language Models (LLMs) beyond static parametric knowledge by enabling access to up-to-date and long-tail information unavaila

Everyone’s quitting but everything’s going great. AGI will be here next week. Scout’s honor!

Everyone’s quitting but everything’s going great. AGI will be here next week. Scout’s honor! This story has now been updated with more details. Three leaders departed from OpenAI today: - Kevin Weil,

safetygary-marcus--x

Exploration and Exploitation Errors Are Measurable for Language Model Agents

arXiv:2604.13151v1 Announce Type: new Abstract: Language Model (LM) agents are increasingly used in complex open-ended decision-making tasks, from AI coding to physical AI. A core requirement in these

safetyarxiv-cs-ai

Filling in the Mechanisms: How do LMs Learn Filler-Gap Dependencies under Developmental Constraints?

arXiv:2604.14459v1 Announce Type: new Abstract: For humans, filler-gap dependencies require a shared representation across different syntactic constructions. Although causal analyses suggest this may

Flow with the Force Field: Learning 3D Compliant Flow Matching Policies from Force and Demonstration-Guided Simulation Data

arXiv:2510.02738v3 Announce Type: replace-cross Abstract: While visuomotor policy has made advancements in recent years, contact-rich tasks still remain a challenge. Robotic manipulation tasks that re

From Boundaries to Semantics: Prompt-Guided Multi-Task Learning for Petrographic Thin-section Segmentation

arXiv:2604.14805v1 Announce Type: new Abstract: Grain-edge segmentation (GES) and lithology semantic segmentation (LSS) are two pivotal tasks for quantifying rock fabric and composition. However, thes

From Plausible to Causal: Counterfactual Semantics for Policy Evaluation in Simulated Online Communities

arXiv:2604.03920v2 Announce Type: replace Abstract: LLM-based social simulations can generate believable community interactions, enabling ``policy wind tunnels'' where governance interventions are tes

GFT: From Imitation to Reward Fine-Tuning with Unbiased Group Advantages and Dynamic Coefficient Rectification

arXiv:2604.14258v1 Announce Type: cross Abstract: Large language models are typically post-trained using supervised fine-tuning (SFT) and reinforcement learning (RL), yet effectively unifying efficien

GSNR: Graph Smooth Null-Space Representation for Inverse Problems

arXiv:2602.20328v2 Announce Type: replace Abstract: Inverse problems in imaging are ill-posed, leading to infinitely many solutions consistent with the measurements due to the non-trivial null-space o

Heat and Matern Kernels on Matchings

arXiv:2604.14331v1 Announce Type: new Abstract: Applying kernel methods to matchings is challenging due to their discrete, non-Euclidean nature. In this paper, we develop a principled framework for co

Hierarchical Retrieval Augmented Generation for Adversarial Technique Annotation in Cyber Threat Intelligence Text

arXiv:2604.14166v1 Announce Type: new Abstract: Mapping Cyber Threat Intelligence (CTI) text to MITRE ATT&CK technique IDs is a critical task for understanding adversary behaviors and automating threa

Hybrid Latents -- Geometry-Appearance-Aware Surfel Splatting

arXiv:2604.14928v1 Announce Type: new Abstract: We introduce a hybrid Gaussian-hash-grid radiance representation for reconstructing 2D Gaussian scene models from multi-view images. Similar to NeST spl

I went on @BBCNewsnight this week to discuss the recent developments in AI's capabilities, as well as the potential harms and concentration …

I went on @BBCNewsnight this week to discuss the recent developments in AI's capabilities, as well as the potential harms and concentration of power they could entail. We need coordinated internationa

safetyyoshua-bengio--x

IG-Search: Step-Level Information Gain Rewards for Search-Augmented Reasoning

arXiv:2604.15148v1 Announce Type: cross Abstract: Reinforcement learning has emerged as an effective paradigm for training large language models to perform search-augmented reasoning. However, existin

Implicit Neural Representations: A Signal Processing Perspective

arXiv:2604.15047v1 Announce Type: new Abstract: Implicit neural representations (INRs) mark a fundamental shift in signal modeling, moving from discrete sampled data to continuous functional represent

Improving Machine Learning Performance with Synthetic Augmentation

arXiv:2604.14498v1 Announce Type: cross Abstract: Synthetic augmentation is increasingly used to mitigate data scarcity in financial machine learning, yet its statistical role remains poorly understoo

in 1996 mitzenmacher showed that sampling two backends and picking the better one drops max load exponentially vs. random selection one extr…

in 1996 mitzenmacher showed that sampling two backends and picking the better one drops max load exponentially vs. random selection one extra comparison. that's the whole trick we built pinecone assis

safetypinecone--x

Language of Thought Shapes Output Diversity in Large Language Models

arXiv:2601.11227v2 Announce Type: replace Abstract: Output diversity is crucial for Large Language Models as it underpins pluralism and creativity. In this work, we reveal that controlling the languag

Layered Mutability: Continuity and Governance in Persistent Self-Modifying Agents

arXiv:2604.14717v1 Announce Type: cross Abstract: Persistent language-model agents increasingly combine tool use, tiered memory, reflective prompting, and runtime adaptation. In such systems, behavior

LeapAlign: Post-Training Flow Matching Models at Any Generation Step by Building Two-Step Trajectories

arXiv:2604.15311v1 Announce Type: new Abstract: This paper focuses on the alignment of flow matching models with human preferences. A promising way is fine-tuning by directly backpropagating reward gr

Learning Ad Hoc Network Dynamics via Graph-Structured World Models

arXiv:2604.14811v1 Announce Type: new Abstract: Ad hoc wireless networks exhibit complex, innate and coupled dynamics: node mobility, energy depletion and topology change that are difficult to model a

Learning Adaptive Reasoning Paths for Efficient Visual Reasoning

arXiv:2604.14568v1 Announce Type: cross Abstract: Visual reasoning models (VRMs) have recently shown strong cross-modal reasoning capabilities by integrating visual perception with language reasoning.

Learning to Think Like a Cartoon Captionist: Incongruity-Resolution Supervision for Multimodal Humor Understanding

arXiv:2604.15210v1 Announce Type: cross Abstract: Humor is one of the few cognitive tasks where getting the reasoning right matters as much as getting the answer right. While recent work evaluates hum

MARS^2: Scaling Multi-Agent Tree Search via Reinforcement Learning for Code Generation

arXiv:2604.14564v1 Announce Type: cross Abstract: Reinforcement learning (RL) paradigms have demonstrated strong performance on reasoning-intensive tasks such as code generation. However, limited traj

Mean Flow Policy Optimization

arXiv:2604.14698v1 Announce Type: new Abstract: Diffusion models have recently emerged as expressive policy representations for online reinforcement learning (RL). However, their iterative generative

Meituan Merchant Business Diagnosis via Policy-Guided Dual-Process User Simulation

arXiv:2604.15190v1 Announce Type: cross Abstract: Simulating group-level user behavior enables scalable counterfactual evaluation of merchant strategies without costly online experiments. However, bui

Metric-Aware Principal Component Analysis (MAPCA):A Unified Framework for Scale-Invariant Representation Learning

arXiv:2604.14249v1 Announce Type: new Abstract: We introduce Metric-Aware Principal Component Analysis (MAPCA), a unified framework for scale-invariant representation learning based on the generalised

Model-Free Assessment of Simulator Fidelity via Quantile Curves

arXiv:2512.05024v3 Announce Type: replace-cross Abstract: As generative AI models are increasingly used to simulate real-world systems, quantifying the ``sim-to-real'' gap is critical. For each input

Modeling LLM Unlearning as an Asymmetric Two-Task Learning Problem

arXiv:2604.14808v1 Announce Type: new Abstract: Machine unlearning for large language models (LLMs) aims to remove targeted knowledge while preserving general capability. In this paper, we recast LLM

Multi-Modal Manipulation via Multi-Modal Policy Consensus

arXiv:2509.23468v3 Announce Type: replace-cross Abstract: Effectively integrating diverse sensory modalities is crucial for robotic manipulation. However, the typical approach of feature concatenation

Multi-Persona Thinking for Bias Mitigation in Large Language Models

arXiv:2601.15488v2 Announce Type: replace Abstract: Large Language Models (LLMs) exhibit social biases, which can lead to harmful stereotypes and unfair outcomes. We propose extbf{Multi-Persona Thinki

Multi-User mmWave Beam and Rate Adaptation via Combinatorial Satisficing Bandits

arXiv:2604.14908v1 Announce Type: new Abstract: We study downlink beam and rate adaptation in a multi-user mmWave MISO system where multiple base stations (BSs), each using analog beamforming from fin

NG-GS: NeRF-Guided 3D Gaussian Splatting Segmentation

arXiv:2604.14706v1 Announce Type: new Abstract: Recent advances in 3D Gaussian Splatting (3DGS) have enabled highly efficient and photorealistic novel view synthesis. However, segmenting objects accur

NLP needs Diversity outside of 'Diversity'

arXiv:2604.14595v1 Announce Type: new Abstract: This position paper argues that recent progress with diversity in NLP is disproportionately concentrated on a small number of areas surrounding fairness

ORBIT: On-policy Exploration-Exploitation for Controllable Multi-Budget Reasoning

arXiv:2601.08310v2 Announce Type: replace Abstract: Recent Large Reasoning Models (LRMs) achieve strong performance by leveraging long-form Chain-of-Thought (CoT) reasoning, but uniformly applying ove

Practical estimation of the optimal classification error with soft labels and calibration

arXiv:2505.20761v3 Announce Type: replace Abstract: While the performance of machine learning systems has experienced significant improvement in recent years, relatively little attention has been paid

Preconditioned Test-Time Adaptation for Out-of-Distribution Debiasing in Narrative Generation

arXiv:2603.13683v2 Announce Type: replace Abstract: Although debiased large language models (LLMs) excel at handling known or low-bias prompts, they often fail on unfamiliar and high-bias prompts. We

PROXIMA: A Reliability Scoring Framework for Proxy Metrics in Online Controlled Experiments

arXiv:2604.14352v1 Announce Type: cross Abstract: Online A/B testing at scale relies on proxy metrics -- short-term, easily-measured signals used in place of slow-moving long-term outcomes. When the p

Pushing the Boundaries of Multiple Choice Evaluation to One Hundred Options

arXiv:2604.14634v1 Announce Type: new Abstract: Multiple choice evaluation is widely used for benchmarking large language models, yet near ceiling accuracy in low option settings can be sustained by s

QU-NLP at ArchEHR-QA 2026: Two-Stage QLoRA Fine-Tuning of Qwen3-4B for Patient-Oriented Clinical Question Answering and Evidence Sentence Alignment

arXiv:2604.14175v1 Announce Type: new Abstract: We present a unified system addressing both Subtask 3 (answer generation) and Subtask 4 (evidence sentence alignment) of the ArchEHR-QA Shared Task. For

R3D: Revisiting 3D Policy Learning

arXiv:2604.15281v1 Announce Type: new Abstract: 3D policy learning promises superior generalization and cross-embodiment transfer, but progress has been hindered by training instabilities and severe o

RaTA-Tool: Retrieval-based Tool Selection with Multimodal Large Language Models

arXiv:2604.14951v1 Announce Type: cross Abstract: Tool learning with foundation models aims to endow AI systems with the ability to invoke external resources -- such as APIs, computational utilities,

Reinforcement Learning via Value Gradient Flow

arXiv:2604.14265v1 Announce Type: new Abstract: We study behavior-regularized reinforcement learning (RL), where regularization toward a reference distribution (the dataset in offline RL or the base m

Reward-Aware Trajectory Shaping for Few-step Visual Generation

arXiv:2604.14910v1 Announce Type: new Abstract: Achieving high-fidelity generation in extremely few sampling steps has long been a central goal of generative modeling. Existing approaches largely rely

RoSLAC: Robust Simultaneous Localization and Calibration of Multiple Magnetometers

arXiv:2604.14353v1 Announce Type: new Abstract: Localization of autonomous mobile robots (AMRs) in enclosed or semi-enclosed environments such as offices, hotels, hospitals, indoor parking facilities,

safetyarxiv-cs-ro