safety

Boundary Suppression Asymmetry in Post-trained Assistants: Over-expansion as a Controllability Cost

arXiv:2605.27969v1 Announce Type: new Abstract: Post-trained language-model assistants are often optimized to avoid under-answering, encouraging complete, helpful, cautious, and proactive responses. W

Safety

BPPO: Binary Prefix Policy Optimization for Efficient GRPO-Style Reasoning RL with Concise Responses

DGX agent

arXiv:2605.28028v1 Announce Type: new Abstract: Group Relative Policy Optimization (GRPO) is widely used for training reasoning models, but updating all sampled completions in each group incurs substa

safetyarxiv-cs-lg

28 May 2026

Breaking the Script Barrier: Enabling Automatic Alignment for PoS-based ASR Error Analysis in Non-Latin Scripts

arXiv:2605.28438v1 Announce Type: new Abstract: Automatic Speech Recognition (ASR) systems are commonly evaluated using aggregate metrics such as Word Error Rate (WER), which do not capture the lingui

Bridging the Detection-to-Abstention Gap in Reasoning Models under Insufficient Information

arXiv:2605.28070v1 Announce Type: new Abstract: We highlight a failure mode of large reasoning models on questions with insufficient information: models may recognize that a problem is under-specified

Calibrated Inference for the Conditional Average Treatment Effect in the Few-Placebo Regime via Gaussian Processes

arXiv:2605.27473v1 Announce Type: cross Abstract: Estimating how much an intervention helps a given individual the conditional average treatment effect (CATE) is increasingly central to decision-makin

CANDOR: Counterfactual ANnotated DOubly Robust Off-Policy Evaluation

arXiv:2412.08052v2 Announce Type: replace Abstract: Off-policy evaluation (OPE) is critical for applying contextual bandit algorithms to high-stakes decision-making settings such as healthcare, where

Causal Direct Preference Optimization for Distributionally Robust Generative Recommendation

arXiv:2603.22335v2 Announce Type: replace-cross Abstract: Direct Preference Optimization (DPO) guides large language models (LLMs) to generate recommendations aligned with user historical behavior dis

Causal Machine Learning: A Survey and Open Problems

arXiv:2206.15475v3 Announce Type: replace Abstract: Causal Machine Learning (CausalML) is an umbrella term for machine learning methods that formalize the data-generation process as a structural causa

Chance-Constrained MPPI under State and Dynamic Object Prediction Uncertainty and the Evaluation of Collision Risk Calibration

arXiv:2605.28330v1 Announce Type: new Abstract: Chance-constrained Model Predictive Path Integral (MPPI) control is increasingly adopted for navigation in dynamic environments to explicitly bound coll

safetyarxiv-cs-ro

CIRF: Tokenizing Chain-of-Thoughts into Reusable Functional Units for Efficient Latent Reasoning in Large Language Models

arXiv:2605.28292v1 Announce Type: new Abstract: Implicit Chain-of-Thought (CoT) reduces the inference cost of large language models by internalizing the explicit rationales. However, existing approach

CodeGENCAT: Generative Computerized Adaptive Testing for Open-ended Coding Problems

arXiv:2602.20020v2 Announce Type: replace Abstract: Existing Computerized Adaptive Testing (CAT) frameworks typically select questions based on the predicted likelihood that the student will answer co

Colosseum: Auditing Collusion in Cooperative Multi-Agent Systems

arXiv:2602.15198v2 Announce Type: replace-cross Abstract: Multi-agent systems, where LLM agents communicate through free-form language, enable sophisticated coordination for solving complex cooperativ

Commit to the Bit: Reactive Reinforcement Learning Done Right

arXiv:2605.28276v1 Announce Type: new Abstract: Reinforcement learning algorithms are commonly analyzed (and designed) under the Markov assumption. This is unrealistic, as most environments encountere

Compositional Text-to-Image Generation Via Region-aware Bimodal Direct Preference Optimization

arXiv:2605.28615v1 Announce Type: new Abstract: Despite the rapid progress of text-to-image (T2I) models, generating images that accurately reflect complex compositional prompts (covering attribute bi

Con-DSO: Learning Short-Horizon Consistency Priors for RGB-D Direct Sparse Odometry

arXiv:2605.27952v1 Announce Type: new Abstract: Visual odometry (VO) is a fundamental component in robotics and augmented reality. RGB-D direct VO benefits from metric depth measurements, but it can d

COTTA: Context-Aware Transfer Adaptation for Trajectory Prediction in Autonomous Driving

arXiv:2604.00402v2 Announce Type: replace-cross Abstract: Developing robust models to accurately predict the trajectories of surrounding agents is fundamental to autonomous driving safety. However, mo

Counterfactually Fair Regression via Optimal Transport

arXiv:2605.28251v1 Announce Type: cross Abstract: We consider the problem of learning a counterfactually fair regressor. We adopt a causal uncertainty view in which counterfactual fairness is defined

CPPO: Contrastive Perception Policy Optimization for VLM Agents

arXiv:2601.00501v2 Announce Type: replace Abstract: We introduce CPPO, a Contrastive Perception Policy Optimization method for finetuning vision--language models (VLMs). Reliable perception is a core

CRaFT: Circuit-Guided Refusal Feature Selection via Cross-Layer Transcoders

arXiv:2604.01604v2 Announce Type: replace Abstract: While modern LLMs are aligned to refuse harmful requests, it is essential to understand the underlying mechanistic basis of this refusal behavior fo

crazy that this was announced on the day tokenmaxxing died.

crazy that this was announced on the day tokenmaxxing died. Anthropic raised 65 billion in a funding round that valued the artificial intelligence company at 965 billion including the new investment,

safetygary-marcus--x

Cross-Entropy Games and Frost Training

arXiv:2605.27701v1 Announce Type: new Abstract: We present Frost Training, a method for improving Monte Carlo-based policy optimization for a large family of LLM-as-a-judge tasks called Cross-Entropy

Cyberbullying Governance on Social Media: A Unified Framework from Content Identification to Intervention

arXiv:2605.27584v1 Announce Type: new Abstract: The proliferation of social media platforms and online communities has inadvertently catalyzed the spread of cyberbullying, hate speech, and other forms

DebFilter: Eradicating Biases Stashed in Value

arXiv:2605.28167v1 Announce Type: new Abstract: Text-to-image diffusion models, which are theoretically equivalent to score-based generative models, generate images through a multi-step denoising proc

Deconstructing Spatial Complexity: Hierarchical Decomposition for LLM Spatial Reasoning

arXiv:2605.28144v1 Announce Type: new Abstract: LLMs have shown remarkable proficiency in general language understanding and reasoning. However, they consistently underperform in spatial reasoning tha

Delay-Aware Reinforcement Learning for Highway On-Ramp Merging under Stochastic Communication Latency

arXiv:2403.11852v5 Announce Type: replace-cross Abstract: Delayed and partially observable state information poses significant challenges for reinforcement learning (RL)-based control in real-world au

DenoiseRL: Bootstrapping Reasoning Models to Recover from Noisy Prefixes

arXiv:2605.28421v1 Announce Type: new Abstract: Reinforcement learning has become a central paradigm for advancing reasoning in large language models, yet most existing methods still depend on stronge

Diagnosing Live Within-Policy Instruction Conflicts in LLM Agents with Witnessed Resolution Profiles

arXiv:2605.27784v1 Announce Type: new Abstract: LLM agents are governed by long-lived natural-language prompt policies, but individually reasonable standing rules can interact in uninspected ways. We

Did AI just solve math? Cal Newport podcast: https://youtu.be/fhZRWZ6J4k4?si=Gzut_5x856n5DW6o

Cal Newport discusses whether recent AI advances represent a breakthrough in mathematical problem-solving capabilities, examining the implications of AI systems' improved performance on complex mathem

safetygary-marcus--x

Differentiable Model Predictive Safety for Heterogeneous Mobility at Urban Intersections

arXiv:2605.27418v1 Announce Type: cross Abstract: The imminent integration of autonomous vehicles and mobile robots in urban settings presents a critical safety challenge for future intelligent transp

safetyarxiv-cs-ro

Diffusion-Augmented Markov Decision Processes for Maximum Entropy Reinforcement Learning

arXiv:2512.02019v3 Announce Type: replace-cross Abstract: Diffusion models excel at sampling from complex, unnormalized distributions. In this work, we extend Maximum Entropy Reinforcement Learning (M

Diffusion Large Language Models for Visual Speech Recognition

arXiv:2605.28456v1 Announce Type: new Abstract: Existing Visual Speech Recognition (VSR) systems commonly rely on left-to-right autoregressive decoding, which can force premature decisions on visually

DiscoForcing: A Unified Framework for Real-Time Audio-Driven Character Control with Diffusion Forcing

arXiv:2605.28491v1 Announce Type: new Abstract: We study real-time audio-responsive character control as a deployment-faithful problem: strictly causal, bounded-latency streaming that must generate co

Disentangling Adversarial Prompts: A Semantic-Graph Defense for Robust LLM Security

arXiv:2605.27823v1 Announce Type: cross Abstract: Large Language Models (LLMs) are increasingly vulnerable to adversarial prompts that exploit semantic ambiguities to bypass safety mechanisms, resulti

DREAM-R: Multimodal Speculative Reasoning with RL-Based Refined Drafting, Precise Verification, and Fully Parallel Execution

arXiv:2605.28678v1 Announce Type: new Abstract: Speculative reasoning has recently been proposed as a means to accelerate reasoning-intensive generation in large multimodal models, but its effectivene

EAPO: Entropy-Driven Adaptive Positive-Negative Sample Weighting for Policy Optimization in Open-Ended QA

arXiv:2605.27846v1 Announce Type: new Abstract: Large Reasoning Models are typically trained via reinforcement learning from verifiable rewards (RLVR). However, existing approaches adopt fixed weights

ECHO: Entropy-Confidence Hybrid Optimization for Test-Time Reinforcement Learning

arXiv:2602.02150v2 Announce Type: replace-cross Abstract: Test-time reinforcement learning generates multiple candidate answers via repeated rollouts and performs online updates using pseudo-labels co

Emerging Extrinsic Dexterity in Cluttered Scenes via Dynamics-aware Policy Learning

arXiv:2603.09882v2 Announce Type: replace-cross Abstract: Extrinsic dexterity leverages environmental contact to overcome the limitations of prehensile manipulation. However, achieving such dexterity

EntroAD: Structural Entropy-Guided Prompt Adaptation for Zero-Shot Anomaly Detection

arXiv:2605.28630v1 Announce Type: new Abstract: Zero-Shot Anomaly Detection (ZSAD) aims to detect anomalies in unseen domains without target-domain adaptation. Recent CLIP-based methods have shown pro

Escape the Language Prior: Mitigating Late-Stage Modality Collapse in Audio Reasoning via Modality-Aware Policy Optimization

arXiv:2605.27741v1 Announce Type: new Abstract: Audio and omni-modal large language models exhibit impressive cross-modal reasoning capabilities. However, applying standard reinforcement learning post

Evaluating the Realism of LLM-powered Social Agents: A Case Study of Reactions to Spanish Online News

arXiv:2605.28598v1 Announce Type: cross Abstract: LLM-powered social agents are increasingly used to simulate online social behavior, yet their realism remains difficult to validate. Existing work has

Evolving and Detecting Multi-Turn Deception using Geometric Signatures

arXiv:2605.27671v1 Announce Type: cross Abstract: Safety defenses for large language models (LLMs) are typically trained and evaluated on single-turn prompts, yet real attacks often unfold as indirect

Examining Agents' Bias Amplification versus Suppression in Multi-Agent Systems

arXiv:2605.28098v1 Announce Type: new Abstract: Multi-agent systems are increasingly deployed to support various tasks where agents interact to achieve individual and collective objectives. Although t

F Scott Fitzgerald, writing about someone who may as well have been Peter Thiel: “They were careless people, Tom and Daisy- they smashed up …

F Scott Fitzgerald, writing about someone who may as well have been Peter Thiel: “They were careless people, Tom and Daisy- they smashed up things and creatures and then retreated back into their mone

safetygary-marcus--x

FABSVer: Faster Training and Better Self-Verification for LLM Mathematical Reasoning

arXiv:2605.28389v1 Announce Type: new Abstract: While large language models have made significant progress in mathematical reasoning, they remain unreliable at judging the correctness of their own sol

FedEHR-Gen: Federated Synthetic Time-Series EHR Generation via Latent Space Alignment and Distribution-Aware Aggregation

arXiv:2605.27892v1 Announce Type: new Abstract: Synthetic Electronic Health Record (EHR) generation provides a promising avenue for data augmentation and cross-hospital modeling in privacy-constrained

From Affect to Complex Behavior: Advancing Multimodal Human-Centered AI at the 10th ABAW Workshop & Competition

arXiv:2605.27451v1 Announce Type: new Abstract: The 10th Affective & Behavior Analysis in-the-Wild (ABAW) Workshop and Competition, held at CVPR 2026, continues to advance research on modelling, analy

From AR to Diffusion: Efficiently Adapting Large Language Models with Strictly Causal and Elastic Horizons

arXiv:2605.27387v1 Announce Type: cross Abstract: Diffusion models promise efficient parallel text generation but rely on bidirectional attention, creating a structural mismatch with pre-trained Autor

From Learning Resources to Competencies: LLM-Based Tagging with Evidence and Graph Constraints

arXiv:2605.28483v1 Announce Type: new Abstract: Linking learning resources to a structured competency framework is key to enabling competency-based search and curriculum analytics in Learning Manageme