safety

Why LLMs rarely payoff—and what I have been saying literally for 7 years—confirmed yet again: LLMs can’t handle the truth. (Nor apparently can my critics, who keep saying I am “always wrong”, when I h

A 25B fund just refused SpaceX at any price. It says the company can't be worth more than 1T, half the $1.8T IPO target, and Musk's 85% control makes it impossible to fix from inside. https://thenextw

arXiv:2605.29975v1 Announce Type: new Abstract: We present a fully convolutional denoising autoencoder (FC-DAE) for denoising two-time intensity-intensity correlation functions (C_2) in X-ray photon c

arXiv:2605.29673v1 Announce Type: cross Abstract: Reconstruction-based inference assigns a class by comparing class-wise reconstruction residuals; Sparse Representation Classification (SRC) is a canon

arXiv:2605.30313v1 Announce Type: new Abstract: Simulation-based RL for contemporary robot control is increasingly organized around GPU-resident simulation: physics, rollout collection, and learning a

arXiv:2605.28824v1 Announce Type: new Abstract: Constructing artificial lexicons that are pronounceable, typologically plausible, and semantically structured remains an open challenge in computational

arXiv:2605.30070v1 Announce Type: cross Abstract: Moving beyond simple scalar rewards toward richer world feedback is a natural path to more scalable RL post-training. On-policy self-distillation (OPS

arXiv:2512.11944v2 Announce Type: replace-cross Abstract: Motion planning for autonomous driving (AD) faces a critical trade-off. While traditional rule-based pipelines offer verifiable safety and int

arXiv:2605.29340v1 Announce Type: new Abstract: In this paper, we discuss question-answer dataset for LLM safety evaluation, with a focus on illegal activities. Specifically, on the basis of manual an

arXiv:2605.29791v1 Announce Type: new Abstract: While Large Language Models (LLMs) can convincingly simulate personas in explicit self-reports, they often deviate in implicit behavioral decisions, rev

arXiv:2605.29458v1 Announce Type: cross Abstract: Accurately simulating the decisions of a specific individual remains challenging for large language models (LLMs), partly because persona information

arXiv:2603.01006v2 Announce Type: replace-cross Abstract: REPresentation Alignment (REPA) improves the training of generative flow models by aligning intermediate hidden states with pretrained teacher

arXiv:2605.29910v1 Announce Type: cross Abstract: Consensus protocols form the backbone of distributed systems and blockchains, where implementation bugs can cause data corruption and financial losses

arXiv:2605.28914v1 Announce Type: cross Abstract: Tool-using language agents turn model decisions into external side effects: they read files, run scripts, call APIs, send messages, and invoke Model C

arXiv:2605.29434v1 Announce Type: cross Abstract: Existing sentence-level watermarking methods enhance robustness to paraphrasing by anchoring watermarks in sentence semantics. However, their prefix-b

arXiv:2605.29139v1 Announce Type: cross Abstract: Federated Conformal RAG (FC-RAG) provides distribution-free coverage for a bandwidth-limited swarm of weak language models, but only at a fixed horizo

arXiv:2605.30031v1 Announce Type: cross Abstract: Large Audio Language Models (LALMs) expand jailbreak risks from token-level prompting to the full speech perception-to-reasoning pipeline, where unsaf

arXiv:2605.29202v1 Announce Type: new Abstract: Recent advances in text-to-music generation enable high-fidelity synthesis of structured musical audio, raising growing concerns about data provenance,

arXiv:2605.30208v1 Announce Type: cross Abstract: AI-assisted coding tools have altered software production. At Meta, significant lines of code per human-landed diff grew by 105.9% year over year and

arXiv:2605.28994v1 Announce Type: new Abstract: AI tools to support real world decision making must be able to build simulation models that inform their recommendations and render them interpretable.

arXiv:2605.28849v1 Announce Type: new Abstract: Gradient temporal-difference methods provide stable off-policy prediction with linear function approximation, but their practical performance is strongl

arXiv:2605.29629v1 Announce Type: new Abstract: Attack Success Rate (ASR) evaluates each jailbreak with a single yes/no label at the end of generation, telling us whether a failure happened but not ho

arXiv:2605.29414v1 Announce Type: cross Abstract: Recent studies have shown that code-switching data (CSD), in which multiple languages are mixed within the same context, can improve cross-lingual tra

arXiv:2605.29697v1 Announce Type: new Abstract: In Agentic Search, trajectory-level outcome rewards fail to quantify the behavioral contributions of individual steps, while existing step-level reward

bold set of counterpredictions, from @scaling01: Cold take on what comes next: - OpenAI will flourish - Anthropic will continue to be profitable - Google will not catch up to Anthropic or OpenAI - no

arXiv:2605.30226v1 Announce Type: cross Abstract: Vision-Language-Action (VLA) models have emerged as a promising paradigm for grounding visual-language understanding into real-world robotic manipulat

arXiv:2605.29078v1 Announce Type: new Abstract: Event-driven scheduling policies are increasingly deployed in industrial environments, where decisions are made under asynchronous and partially observe

arXiv:2601.08064v2 Announce Type: replace Abstract: Confidence estimation (CE) indicates how reliable the answers of large language models are and impacts user trust and decision-making. Existing eval

arXiv:2605.29971v1 Announce Type: new Abstract: Causal interventions in language model representations have largely targeted discrete features, like grammatical number. However, language models must a

arXiv:2602.11389v2 Announce Type: replace Abstract: World models require robust relational understanding to support prediction, reasoning, and control. While object-centric representations provide a u

arXiv:2605.29836v1 Announce Type: cross Abstract: Despite strong average-case performance, deep learning models often exhibit systematic errors on specific population groups, known as error slices. Id

arXiv:2605.29788v1 Announce Type: new Abstract: Critical sequential decisions are rarely single-timescale: a strategic decision causally shapes the context in which every subsequent tactical choice is

arXiv:2605.30332v1 Announce Type: new Abstract: Diffusion models achieve state-of-the-art image synthesis, with their generative trajectories fundamentally exhibiting a spectral bias, resolving low-fr

arXiv:2605.29452v1 Announce Type: new Abstract: Image-based 3D reconstruction offers a low-cost alternative to traditional sensor-based techniques for road surface assessment. This study compares four

arXiv:2410.07287v2 Announce Type: replace-cross Abstract: Climate change poses an existential threat, necessitating effective climate policies to enact impactful change. Decisions in this domain are i

arXiv:2605.29886v1 Announce Type: cross Abstract: Retrieval-augmented generation (RAG) improves knowledge-intensive question answering by incorporating external evidence. However, existing RAG methods

arXiv:2605.30211v1 Announce Type: new Abstract: Self-supervised video Object-Centric Learning (OCL) aims to discover distinct objects and associate them across time, whereas self-supervised Multi-Obje

arXiv:2605.30135v1 Announce Type: cross Abstract: Various algorithms have been proposed to address the challenges posed by class-imbalanced learning from real-world data with long-tailed distributions

arXiv:2605.29522v1 Announce Type: new Abstract: As scientific literature grows rapidly, automated survey generation has become a key capability for AI scientists and human researchers. However, existi

arXiv:2605.29570v1 Announce Type: new Abstract: Purpose: Laparoscopic ultrasound (LUS) enhances the safety of liver surgery by visualizing intrahepatic vessels in real-time. Still, vessel identificati

arXiv:2605.30215v1 Announce Type: new Abstract: Recent feed-forward 3D reconstruction transformers have scaled to over a billion parameters, following the broader trend of increasing model capacity in

arXiv:2605.30003v1 Announce Type: cross Abstract: We study two-level autoresearch for cooperation: an outer-loop AI agent autonomously redesigns the inner-loop pipeline of an LLM policy-synthesis syst

arXiv:2605.29626v1 Announce Type: cross Abstract: Steering language model generation toward desired textual properties is essential for practical deployment, and inference-time methods are particularl

arXiv:2605.29152v1 Announce Type: new Abstract: Randomly initialized neural networks induce a prior over functions, but the predictor used in practice is produced only after training. We ask how much

arXiv:2605.29343v1 Announce Type: new Abstract: Speculative decoding accelerates large language model inference by pairing a target model with a lightweight draft model whose proposed tokens are verif

arXiv:2510.27607v3 Announce Type: replace Abstract: Augmenting vision-language-action models (VLAs) with world models is promising for robotic policy learning but faces challenges in jointly predictin