Search: “safety”

arXiv:2503.18893v2 Announce Type: replace Abstract: Long-context Large Language Models (LLMs) enable powerful applications but incur high memory costs due to the key-value states (KV-Cache). Recent st

you literally can’t prove or disprove the idea that beaches are conscious so should we stop walking on them? @dash_eats You literally cannot prove or disprove this. Because we cannot prove or disprove

arXiv:2605.26174v1 Announce Type: cross Abstract: Production language-model systems answer a request by partitioning it across an invisible orchestration of worker agents that recompose one integrated

a16z waking up and realizing we are nowhere near the singularity 🤣 OpenAI and Anthropic are effectively telling the market they can't solve every problem with a generic AI coworker. You don't pour bil

arXiv:2605.27095v1 Announce Type: new Abstract: Learning from demonstrations in embodied control is often cast as behavioral cloning, and recent diffusion or flow-matching policies improve this paradi

arXiv:2511.07667v2 Announce Type: replace Abstract: The equitable assessment of individual contribution in teams remains a persistent challenge, where conflict and disparity in workload can result in

arXiv:2605.25763v2 Announce Type: replace Abstract: Text-to-image synthesis has made significant progress, benefiting from the strong generative capabilities of diffusion models. However, these models

arXiv:2605.26130v1 Announce Type: new Abstract: Operational weather prediction at kilometer scales remains computationally prohibitive for traditional numerical weather prediction (NWP) models, limiti

arXiv:2511.16870v3 Announce Type: replace Abstract: Enforcing alignment between the internal representations of diffusion or flow-based generative models and those of pretrained self-supervised encode

arXiv:2506.23149v2 Announce Type: replace Abstract: Reusable skills play a key role in improving LLM-based agents, but existing skill-evolution methods often fail to ensure that evolved skills both co

arXiv:2605.26552v1 Announce Type: cross Abstract: Aligning a few-step generative model is challenging, since existing alignment frameworks typically rely on restrictive assumptions: a tractable likeli

arXiv:2603.17218v2 Announce Type: replace-cross Abstract: Post-training alignment optimizes language models to match human preference signals, but this objective is not equivalent to modeling observed

arXiv:2605.27355v1 Announce Type: new Abstract: Reinforcement Learning from Human Feedback (RLHF) is the standard method to align Large Language Models (LLMs) with human preferences. In this work, we

arXiv:2605.26442v1 Announce Type: cross Abstract: Much of the alignment tuning literature is organized around optimization objectives, while the construction of alignment data is often treated implici

arXiv:2605.26397v1 Announce Type: cross Abstract: Large language models (LLMs) are increasingly used in decision-making tasks where they can amplify or suppress perspectives, raising concerns in high-

arXiv:2605.26434v1 Announce Type: cross Abstract: EEG foundation models, pre-trained on large-scale unlabelled EEG data, have emerged as a promising direction towards learning generalizable EEG repres

arXiv:2601.05028v2 Announce Type: replace Abstract: Equivariance is a powerful inductive bias in neural networks, improving generalisation and physical consistency. Recently, however, non-equivariant

arXiv:2506.09532v5 Announce Type: replace-cross Abstract: We present Athena-PRM, a multimodal process reward model (PRM) designed to evaluate the reward score for each step in solving complex reasonin

arXiv:2605.26559v1 Announce Type: cross Abstract: Tabular foundation models achieve strong accuracy on choice prediction tasks, but their predictions often violate the economic logic those tasks requi

arXiv:2605.27110v1 Announce Type: cross Abstract: In this work, we propose BAIT (Boundary-Aware Iterative Trap), a three-step jailbreak framework that approaches malicious goals through internal discl

arXiv:2605.27293v1 Announce Type: new Abstract: Reinforcement learning with verifiable rewards has become a standard recipe for improving the reasoning abilities of large language models. Existing alg

arXiv:2603.03585v2 Announce Type: replace-cross Abstract: Misinformation is a growing societal threat, and susceptibility to misinformative claims varies across demographic groups due to differences i

arXiv:2601.03525v3 Announce Type: replace-cross Abstract: Effective reward design is a central challenge in Reinforcement Learning (RL) for code generation. Mainstream test-suite-level outcome rewards

arXiv:2605.26491v1 Announce Type: cross Abstract: Preference optimization has emerged as an efficient alternative to online reinforcement learning from human feedback (RLHF) for aligning text-to-image

arXiv:2605.27131v1 Announce Type: cross Abstract: Enterprise data platforms face an enduring tension between domain self-service and holistic governance. The data mesh paradigm proposed decentralized

arXiv:2605.26684v1 Announce Type: cross Abstract: Group-based reinforcement learning (RL) methods have achieved remarkable success in improving the performance of large language models (LLMs) and have

arXiv:2605.26654v1 Announce Type: cross Abstract: Reinforcement learning (RL) often has a hierarchical structure, where an upper-level (UL) learner selects model parameters and a lower-level (LL) deci

arXiv:2605.26182v1 Announce Type: new Abstract: Generating physically buildable brick structures from 3D shapes requires more than geometric reconstruction: the output must also satisfy discrete part

arXiv:2511.14075v2 Announce Type: replace-cross Abstract: Classifier free guidance is a standard method for conditional sampling in diffusion models, but its sampling rule is not aligned with the obje

Completely agree, @hwchase17 ! This is the meta-level breakthrough we've all been waiting for. Self-optimizing loops finally feel production-ready because LangSmith Engine turns evaluation from a manu

arXiv:2605.26657v1 Announce Type: new Abstract: Long-horizon decision problems with cumulative damage couple locally attractive actions to globally adverse outcomes. We identify two orthogonal failure

arXiv:2605.26990v1 Announce Type: cross Abstract: Bayesian experimental design (BED) is a principled framework for data-efficient design of sequential experiments. However, existing BED methods are un

arXiv:2605.26440v1 Announce Type: new Abstract: The rapid advancement of Large Language Models (LLMs) has outpaced the scalability of traditional evaluation benchmarks, which remain heavily dependent

arXiv:2605.27076v1 Announce Type: cross Abstract: In many multi-agent applications, tasks yield rewards only when executed by a coalition meeting an unknown size threshold; otherwise, feedback is full

arXiv:2605.27115v1 Announce Type: new Abstract: Domain specialization can improve LLM behavior in vertical domains, but often weakens the general capabilities inherited from the original model. Recent

arXiv:2603.21563v2 Announce Type: replace Abstract: Collaborative multi-agent large language models (LLMs) can solve complex reasoning tasks by decomposing roles, but reinforcement learning for such s

arXiv:2605.26385v1 Announce Type: cross Abstract: Large-scale search, recommendation, and retrieval-augmented generation (RAG) systems typically employ a two-stage architecture: an early-stage ranker

arXiv:2605.26293v1 Announce Type: cross Abstract: Prior work establishes that controlled contrastiveness between self-generated responses from large language models, set via reward scores, improves do

arXiv:2510.09405v2 Announce Type: replace Abstract: Radio frequency fingerprint identification (RFFI) is a key technique for wireless network security, leveraging intrinsic hardware imperfections to e

arXiv:2605.26262v1 Announce Type: new Abstract: Museums are important sites for the dissemination of culture and art. They are institutions rooted in history and tradition; their exhibitions are often

arXiv:2605.26236v1 Announce Type: new Abstract: Co-speech gesture generation requires both semantic expressivity and biomechanically plausible rhythmic motion. Existing holistic gesture models mix lex

arXiv:2605.26656v1 Announce Type: new Abstract: Multimodal large language models are typically trained end-to-end to predict ground-truth answers, yet supervision signals are applied exclusively to te

arXiv:2605.26952v1 Announce Type: new Abstract: Agentic reinforcement learning (RL) has proven effective for training LLM-based agents with external tool-use capabilities. However, we identify that ag

arXiv:2605.26478v1 Announce Type: cross Abstract: We present the stochastic decoupled policy gradient (SDPG), a lightweight visual reinforcement learning (RL) method that trains diverse visuomotor con

arXiv:2605.26492v1 Announce Type: cross Abstract: LLM-generated stories are a popular use case, but they show very low variability. We sample 20,000 total stories from four current models using five p

Elon Musk on why humanity must become multiplanetary: “I think it's important for the long-term preservation and ultimately the expansion and extension of the scope and scale of consciousness... that