safety

arXiv:2605.19830v1 Announce Type: new Abstract: Conventional treatment policies map patient covariates to a single recommended intervention in order to maximize expected clinical outcomes. Although a

arXiv:2605.19219v1 Announce Type: new Abstract: A/B testing remains the gold standard for evaluating modifications to e-commerce storefronts, yet it diverts traffic, requires weeks to reach statistica

arXiv:2605.20069v1 Announce Type: new Abstract: Competitive selection processes, from scientific funding to admissions and hiring, use evaluations to score candidates, and eventually choose a subset o

arXiv:2605.20085v1 Announce Type: new Abstract: Robotic manipulation is often specified through language instructions or task identifiers, yet cluttered environments with similar objects are better ha

arXiv:2605.19804v1 Announce Type: cross Abstract: For practical use, diffusion- or flow-based generative models must be aligned with task-specific rewards, such as prompt fidelity or aesthetic prefere

arXiv:2605.19876v1 Announce Type: new Abstract: Text-to-3D generation based on diffusion models often suffers from the Janus problem, leading to inconsistent geometry across viewpoints. This work iden

arXiv:2605.19931v1 Announce Type: cross Abstract: Estimating forest aboveground biomass (AGB) from Earth observation combines two structurally incompatible label sources: spaceborne lidar provides can

arXiv:2605.18988v1 Announce Type: cross Abstract: The expansion of Multimodal Large Language Models (MLLMs) and their integration into autonomous agentic workflows has introduced a non-stationary atta

arXiv:2605.19264v1 Announce Type: new Abstract: Voting methods weighted by stakes are the fundamental governance paradigm in Proof-of-Stake (PoS) blockchains. Such a paradigm is known to be prone to p

arXiv:2605.18816v1 Announce Type: cross Abstract: Neural surrogates enable orders-of-magnitude acceleration of computational fluid dynamics (CFD) simulations, with the potential to transform engineeri

arXiv:2603.07018v2 Announce Type: replace-cross Abstract: Treatment effects estimated from a randomized controlled trial are local not only to the study population but also to the time at which the tr

arXiv:2605.18843v1 Announce Type: new Abstract: Backtesting large language models on historical events requires reasoning exclusively from information available before a specified cutoff date. Yet mod

arXiv:2605.20066v1 Announce Type: new Abstract: Knowledge graph question answering seeks to translate natural language questions into executable queries over knowledge graphs, but existing approaches

arXiv:2605.19638v1 Announce Type: cross Abstract: As large language models (LLMs) demonstrate increasing competence in synthesizing functional user interfaces, a fundamental question emerges in access

The consensus of analysts is that AI capital investments will rise by 20% a year for five years while revenues are expected to grow 15% resulting in negative returns. 'The AI boom will become a story

the crazy part is that people are “clowning” me without knowing anything about the training or whether anything else other than scaled changed or how the model does on anything else. (or what it costs

This video from the World Science Festival, shared by cognitive scientist Gary Marcus, examines the gap between AI systems' claimed reasoning capabilities and their actual mechanisms, likely arguing t

think you know @garymarcus because read a few of his tweets? try watching this, to see the real deal, and why, for example, the US Senate invited him to testify. Superb conversation between @bgreene a

arXiv:2605.20087v1 Announce Type: cross Abstract: Conversational AI has now reached billions of users, yet existing datasets capture only what people say, not what they think. We introduce ThoughtTrac

arXiv:2605.19064v1 Announce Type: cross Abstract: Workforce transformations are difficult to forecast and costly to mismanage. In particular, the integration of artificial intelligence into knowledge

arXiv:2605.20074v1 Announce Type: new Abstract: Distillation transfers knowledge from a large model trained on broad data to a smaller, more efficient model suitable for deployment. In structured pred

arXiv:2605.19035v1 Announce Type: new Abstract: The rapid advancement of Large Language Models has given rise to autonomous LLM-based agents capable of complex reasoning and execution. As these agents

arXiv:2602.11767v3 Announce Type: replace Abstract: Advances in large language models (LLMs) are driving a shift toward using reinforcement learning (RL) to train agents from iterative, multi-turn int

two trillion dollars to build “pathologically dishonest” AI I think @RyanPGreenblatt's recent post summarizes the vibe of this behavior well. Not the sort of thing that would be acceptable for a human

arXiv:2603.18003v4 Announce Type: replace Abstract: Multimodal large language models (MLLMs) exhibit strong visual-language reasoning, yet cannot process structured, non-visual data such as human skel

arXiv:2605.20037v1 Announce Type: cross Abstract: Reward-poisoning attacks present a significant risk to learning-based wireless control systems. Given this, we propose a Disagreement-Guided Reward Po

arXiv:2605.19839v1 Announce Type: new Abstract: Preference alignment aims to guide generative models by learning from comparisons between preferred and non-preferred samples. In practice, most existin

arXiv:2605.19662v1 Announce Type: new Abstract: Tabular foundation models based on pretrained prior-data fitted networks~(PFNs) have shown strong generalization on diverse tabular tasks, but they are

arXiv:2605.19425v1 Announce Type: cross Abstract: Reinforcement Learning with Verifiable Rewards (RLVR) has become the dominant paradigm for advanced reasoning in Large Language Models (LLMs), but rol

arXiv:2602.07008v2 Announce Type: replace Abstract: Reliable models should not only predict correctly, but also justify decisions with acceptable evidence. Yet conventional supervised learning typical

arXiv:2605.19214v1 Announce Type: cross Abstract: Diagnostic performance in medical AI varies systematically across demographic groups, yet subgroup AUC can mask clinically important disparities. At a

arXiv:2605.16913v1 Announce Type: cross Abstract: Neural networks trained with gradient-based methods exhibit a strong simplicity bias: they learn simpler statistical features of their data before mov

arXiv:2605.18224v1 Announce Type: cross Abstract: This note studies exact constant collapse in variational autoencoders, where the encoder mean becomes independent of the input. The goal is to make th

A large-scale meta-analysis examining over 361,000 job applications across nearly 30 countries spanning 40 years found that hiring discrimination based on protected characteristics (such as race, gend

arXiv:2504.14820v2 Announce Type: replace Abstract: For peg-in-hole tasks, humans rely on binocular visual perception to locate the peg above the hole surface and then proceed with insertion. This pap

arXiv:2605.17413v1 Announce Type: cross Abstract: Safety-aligned language models often refuse cybersecurity requests whose wording resembles misuse, even when the task is authorized and defensive. Thi

arXiv:2605.18743v1 Announce Type: new Abstract: Inspired by the emergent behaviors in large language models that generalized human intelligence, the research community is pursuing similar emergent cap

arXiv:2510.04309v3 Announce Type: replace Abstract: Controlling the behaviors of large language models (LLM) is fundamental to their safety alignment and reliable deployment. However, existing steerin

arXiv:2602.02236v4 Announce Type: replace-cross Abstract: We study online fine-tuning of pretrained control policies for autonomous driving using Real-Time Recurrent Reinforcement Learning (RTRRL), a

arXiv:2605.18459v1 Announce Type: new Abstract: Adaptive experimentation enables efficient estimation of causal effects, but existing methods are not designed for survival data with censoring, where e

arXiv:2605.17609v1 Announce Type: new Abstract: Many inference-time language-model pipelines combine a cheap reward signal with an expensive verifier, such as exact answer checking in mathematical rea

arXiv:2605.16993v1 Announce Type: cross Abstract: Current clinical artificial intelligence (AI) systems are evaluated almost exclusively on clean, standardised, English-language inputs, conditions tha

arXiv:2605.17517v1 Announce Type: new Abstract: Recent advances in Vision-Language-Action (VLA) models have shown strong potential for general-purpose robotic manipulation. However, the visual represe

arXiv:2605.17698v1 Announce Type: new Abstract: The deployment of Large Language Models (LLMs) as autonomous economic agents introduces systemic risks that extend beyond individual capability failures

arXiv:2605.17634v1 Announce Type: cross Abstract: Prompt injection is the most critical vulnerability in deployed AI agents. Despite recent progress, we show that the prevailing defense paradigm (data

arXiv:2602.20042v2 Announce Type: replace Abstract: General Alignment has improved average-case helpfulness and safety, but current alignment practice still rewards confident, single-turn responses. T