safety

From Brute Force to Semantic Insight: Performance-Guided Data Transformation Design with LLMs

arXiv:2601.03808v2 Announce Type: replace Abstract: Large language models (LLMs) have achieved notable performance in code synthesis; however, data-aware augmentation remains a limiting factor, handle

Gary Marcus won! @GaryMarcus

DGX agent

Gary Marcus won! @GaryMarcus I would have assumed it was fairly obvious, but in case it's not: a million-line codebase (also known as a 'harness'), running at inference time, orchestrating thousands o

safetygary-marcus--x

6 Aug 2026

Generative Optimization for Incentivized Advertising with Global Level Constraints

arXiv:2608.04421v1 Announce Type: cross Abstract: Incentivized advertising allocates monetary or virtual rewards to drive user engagement, where a key challenge is optimizing continuous incentive magn

GeoReward: Mitigating Contextual Variable Overestimation in Vision-Language Models for Cross-Market Preference Prediction

arXiv:2608.04504v1 Announce Type: cross Abstract: Vision-language models excel in many multimodal tasks but remain prone to a subtle yet impactful failure mode: they tend to overestimate dominant visu

GFlowNet Training by Policy Gradients

arXiv:2408.05885v3 Announce Type: replace Abstract: Generative Flow Networks (GFlowNets) have been shown effective to generate combinatorial objects with desired properties. We here propose a new GFlo

Governing Execution Risk in Agentic AI Systems: A Trajectory-Guided Framework for Red Teaming

arXiv:2608.04018v1 Announce Type: cross Abstract: AI agents are increasingly embedded in organizational workflows, where they interact with external information sources and invoke digital tools to per

Guideline-as-Oracle: Zero-Annotation Training of an Ophthalmic Telephone Triage Agent

arXiv:2608.04772v1 Announce Type: cross Abstract: Scaling supervision for multi-turn medical agents is difficult because expert dialogue annotation is costly and clinical conversations are privacy-res

HALT: Verification-Aware Stopping for Retrieval-Augmented Search Agents

arXiv:2608.02009v2 Announce Type: replace Abstract: Retrieval-augmented search agents answer multi-hop questions by repeatedly issuing search queries and accumulating evidence. This creates a stopping

HCRide: Harmonizing Passenger Fairness and Driver Preference for Human-Centered Ride-Hailing

arXiv:2508.04811v2 Announce Type: replace Abstract: Order dispatch systems play a vital role in ride-hailing services, which directly influence operator revenue, driver profit, and passenger experienc

Integrated Noise and Safety Management in UAM via A Unified Reinforcement Learning Framework

arXiv:2508.16440v2 Announce Type: replace-cross Abstract: Urban Air Mobility (UAM) envisions the widespread use of small aerial vehicles to transform transportation in dense urban environments. Howeve

Interpreting GFlowNets for Drug Discovery: What probes can and cannot show

arXiv:2511.19264v2 Announce Type: replace-cross Abstract: Generative Flow Networks (GFlowNets) construct molecules through sequential decisions, but their internal policies remain opaque, limiting ado

It was the verification problem all along, while the masses were distracted by the alignment problem. Recursive self improvement? How does t…

It was the verification problem all along, while the masses were distracted by the alignment problem. Recursive self improvement? How does the observer observe itself and know that it changed for the

safetyyann-lecun--x

Joint UAV Flight and Opportunistic Routing under Reinforcement Learning for Delay-Tolerant Networks

arXiv:2608.04590v1 Announce Type: new Abstract: The growing deployment of delay-tolerant networks (DTNs) has made store-carry-forward (SCF) communication indispensable under sparse connectivity. Howev

Language Models Generalize to Human-like Word Order Preferences

arXiv:2608.05028v1 Announce Type: new Abstract: A central question in language acquisition is whether linguistic biases can emerge from general learning mechanisms operating over underdetermined input

Learning to Resolve Neutron Resonances with Fully Convolutional Neural Networks

arXiv:2608.04027v1 Announce Type: new Abstract: This work investigates the feasibility of augmenting traditional R-Matrix codes with a robust machine learning framework for automatically detecting neu

Learning When to Stop: Prefix-Optimal Dynamic Diffusion Policies for Continuous Control

arXiv:2608.05084v1 Announce Type: new Abstract: Diffusion policies are a powerful policy class for continuous control, but their iterative denoising process creates a substantial computational bottlen

Lesion Detection in CT with Frozen Self-Distilled Features: SALT, a Spatially Adaptive Label-Guided Temperature

arXiv:2608.05100v1 Announce Type: new Abstract: Self-supervised pretraining objectives are spatially uniform: the teacher temperature and the per-patch loss weight are identical everywhere in the imag

LLMs Struggle to Measure What Distinguishes Students of Different Proficiency Levels: A Study of Item Discrimination in Reading Comprehension Assessment

arXiv:2606.18709v2 Announce Type: replace Abstract: Existing work on LLM-based educational assessment has focused largely on item difficulty, but difficulty alone does not indicate whether an item mea

Local Violation Certification for Linear Predict-Then-Optimize Pipelines

arXiv:2608.04474v1 Announce Type: new Abstract: Data-driven decision pipelines combining predictive machine learning models with downstream optimization software are increasingly used to make high-sta

Long-term Measurements: Towards a Longitudinal Understanding of Human-AI Interactions

arXiv:2608.02491v2 Announce Type: replace Abstract: Language models have taken on the role of a very new type of technology, by virtue of their 'human-ness' and rapid integration into users' daily liv

Looking in the Mirror: Introspecting Side-Effect Misalignments Induced by Fine-Tuning

arXiv:2608.04347v1 Announce Type: new Abstract: Fine-tuning enables a source model to acquire desired capabilities and behaviors in a target domain while retaining much of its general-purpose competen

Manipulation-Proof Oblivious Audits against Deceptive Model Providers

arXiv:2608.04365v1 Announce Type: new Abstract: Audits have emerged as a critical instrument for algorithmic governance, providing a mechanism for external scrutiny and governance of machine learning

Memorization in Large Language Models in Medicine: Prevalence, Characteristics, and Implications

arXiv:2509.08604v5 Announce Type: replace-cross Abstract: Large Language Models (LLMs) have demonstrated significant potential in medicine, with many studies adapting them through continued pre-traini

MGSB: Manifold Gated Signature Branch Pressure-Domain Baseline Architecture for Two-Phase Pipeline Flows Under Distributional Shift

arXiv:2608.04805v1 Announce Type: new Abstract: Leak detection models for multiphase pipelines often degrade when deployed under flow regimes that differ from training. Existing evaluations typically

Multi-Objective Ranking for Live-Streaming: Balancing Fresh and Delayed Signals with Segment-Aware Targeting

arXiv:2608.04455v1 Announce Type: cross Abstract: One of the most challenging problems entertainment live-streaming services face in recommendation systems is that user behaviors are sparse and delaye

Multicalibration Yields Better Matchings

arXiv:2511.11413v2 Announce Type: replace Abstract: Consider the problem of finding the best matching in a weighted graph where we only have access to predictions of the actual stochastic weights, bas

Multimodal Alignment Through Joint Kernel Entropic Gromov--Wasserstein Optimal Transport

arXiv:2608.04234v1 Announce Type: cross Abstract: We study the problem of aligning data from multiple modalities into a shared representation space, focusing on settings where strong pretrained unimod

muSync-GS: Physics-Synchronized Driving Video Synthesis for Weather and Geometric Road Hazards

arXiv:2608.04412v1 Announce Type: new Abstract: High-quality driving data are essential for autonomous-driving systems and generative world models. However, rare and safety-critical scenarios involvin

Non-Stationary Inventory Control with Lead Times

arXiv:2602.05799v2 Announce Type: replace-cross Abstract: We study non-stationary single-item, periodic-review inventory control problems in which the demand distribution is unknown and may change ove

Not All Redundant Tokens Are Alike: Analyzing Visual Token Pruning through Token Roles

arXiv:2608.04483v1 Announce Type: new Abstract: Vision-language models (VLMs) process an image as a sequence of visual tokens, which creates a substantial computational bottleneck during inference. Re

Not Every Divergence Should Be Suppressed: Counterfactual Recoverability in On-Policy Distillation

arXiv:2608.04408v1 Announce Type: cross Abstract: On-policy distillation (OPD) supervises student-visited trajectories, yet divergence-based rules cannot determine whether an erroneous prefix remains

NSF-HRPT: Neural Semantic Field meets Hierarchical Risk Perception Tree for Safety-Critical Scenario Assessment

arXiv:2608.04776v1 Announce Type: new Abstract: The ability to accurately assess and anticipate risks in safety-critical scenarios is crucial for autonomous driving systems. While existing research ha

ODRA: Synthesizing Cognitive Behavioral Therapy Sessions with Structured Chain-Of-Thought and Dynamic Patient Resistance

arXiv:2608.04524v1 Announce Type: new Abstract: Synthetic generation of Cognitive Behavioral Therapy (CBT) sessions is challenged by two competing demands: adhering to strict therapeutic structure whi

OMG i wrote some of the original work on what is to be neurosymbolic in 2001 and dude who probably hasn’t read that work is trying to school…

OMG i wrote some of the original work on what is to be neurosymbolic in 2001 and dude who probably hasn’t read that work is trying to school me on the definition 🤦‍♂️ coding harness and tools calls ar

safetygary-marcus--x

On July 24, the Black at Mila community hosted Prof. Vukosi Marivate @vukosi (University of Pretoria, co-founder of Masakhane, Lelapa AI, an…

On July 24, the Black at Mila community hosted Prof. Vukosi Marivate @vukosi (University of Pretoria, co-founder of Masakhane, Lelapa AI, and Deep Learning Indaba) for a powerful session on the future

safetyyoshua-bengio--x

OPD-V: Visual On-Policy Self-Distillation with Modality Balance

arXiv:2608.05131v1 Announce Type: cross Abstract: On-Policy Self-Distillation (OPSD) has become a standard post-training approach for improving visual reasoning in multimodal large language models (ML

Optimal Constrained sc-LTL Planning in MDPs via Switching Policies

arXiv:2608.05021v1 Announce Type: new Abstract: We study the synthesis of optimal policies for planning problems on Markov decision processes with both objectives and safety constraints specified in c

safetyarxiv-cs-ro

Optimizing What Policies Learn From: Recoverability-aware Rollout Intervention Learning

arXiv:2608.05080v1 Announce Type: cross Abstract: Critic-free group-based reinforcement learning has become a scalable approach for post-training large language models. However, most existing methods

OutLangSplat: 3D Language Gaussian Splatting for UAV Outdoor Scenes

arXiv:2608.04560v1 Announce Type: new Abstract: 3D Language Gaussian Splatting embeds open-vocabulary language features into 3D Gaussian Splatting, providing an efficient explicit representation for t

Overcoming Statistical Bias in Action-Controllable World Models

arXiv:2608.04653v1 Announce Type: new Abstract: Action-conditioned world models aim to predict how visual environments evolve under an agent's actions. Yet future frames are often highly predictable f

Poly-OPD: Heterogeneous Multi-Teacher On-Policy Distillation for Capability-Selectable Flow Models

arXiv:2608.04349v1 Announce Type: new Abstract: Leading open text-to-image models often carry complementary strengths: one may lead on preference-aligned aesthetics while another follows compositional

PRIMAL3: Pathfinding via Reinforcement and Imitation Multi-Agent Learning - Leveraging LaCAM3

arXiv:2608.04905v1 Announce Type: new Abstract: We present PRIMAL3, an ultra-large-scale learning-based framework for multi-agent pathfinding (MAPF) that integrates reinforcement learning, topology-aw

safetyarxiv-cs-ro

Privileged, but Biased: How PI-Conditioned Teachers Break Self-Distillation

arXiv:2608.04794v1 Announce Type: new Abstract: Self-distillation (SD) has emerged as a compute-efficient alternative to reinforcement learning with verifiable rewards: a self-teacher, conditioned on

Q-CueGraph: Query-Conditioned Visual Evidence Graphs for Multimodal Reasoning

arXiv:2608.04452v1 Announce Type: cross Abstract: High-resolution pixels and crop or zoom tools give multimodal large language models the ability to inspect an image, but they do not provide a reliabl

Real-time probabilistic tsunami forecasting via generative AI

arXiv:2608.04327v1 Announce Type: new Abstract: Explicit onshore tsunami inundation forecasting can improve public risk awareness, but deterministically predicted inundation boundaries under highly un

Reward Structure Shapes the Interaction Between Episodic Exploration and Neural Memory in Reinforcement Learning

arXiv:2608.05111v1 Announce Type: new Abstract: In partially observable reinforcement learning, agents face a dual bottleneck: they must explore to encounter rewarding states and retain that experienc

SafeCommit: Certifying When Memory-Grounded Agents May Safely Act

arXiv:2608.04289v1 Announce Type: new Abstract: Long-horizon agents increasingly use persistent memory and tools to take actions with external side effects. A central failure mode is premature commitm

SafeLand: Safe Autonomous Landing in Unknown Environments with Bayesian Semantic Mapping

arXiv:2603.17430v2 Announce Type: replace Abstract: Autonomous landing of uncrewed aerial vehicles (UAVs) in unknown, dynamic environments poses significant safety challenges, particularly near people

safetyarxiv-cs-ro