Knowledge catalogue
arXiv:2608.18091v1 Announce Type: cross Abstract: As LLM-as-a-judge systems become increasingly widespread, self-preference in LLMs -- the tendency to favor one's own outputs -- raises growing concern
arXiv:2608.18263v1 Announce Type: new Abstract: Traffic signal control is a complex sequential decision-making problem requiring real-time adaptation and trade-offs among throughput, delay fairness, s
arXiv:2608.18915v1 Announce Type: new Abstract: Hardware shifts, color variations, and changing patient characteristics between development and deployment routinely break trained medical image classif
arXiv:2608.18852v1 Announce Type: new Abstract: Agent frameworks increasingly package procedural knowledge as skills: instruction files an agent reads on demand, while public libraries now hold thousa
arXiv:2603.04448v2 Announce Type: replace Abstract: Current AI agents can flexibly invoke tools and execute complex tasks, yet their long-term advancement is hindered by the lack of systematic accumul
Rachel Metz / Bloomberg: Source: Anthropic plans a safety system this year that still requires enterprises to retain data for 30 days but lets them do so on their own cloud systems — Anthropic PBC pla
arXiv:2608.19080v1 Announce Type: new Abstract: Object detectors often produce over-confident predictions for objects outside their training categories, leading to so-called out-of-distribution (OoD)
arXiv:2608.18087v1 Announce Type: cross Abstract: Existing subword tokenizers optimize statistical compression but ignore morphological structure, particularly the relationship between roots and affix
arXiv:2608.18715v1 Announce Type: cross Abstract: Ensuring not only high accuracy but also reliable and robust predictions is critical for the deployment of semantic segmentation models in safety-crit
arXiv:2608.18415v1 Announce Type: new Abstract: The edge of stability refers to a phenomenon in deep learning with gradient-based optimizers where the Hessian eigenvalues of the loss remain stable abo
arXiv:2608.18222v1 Announce Type: cross Abstract: Recurrent-depth reasoners aim to solve harder problems by iterating their update longer at test time, but additional iterations can improve, preserve,
this has always been OpenAI’s game plan. (I first warned of it in 2024, saying that surveillance might be their only viable business plane.) 🦔ChatGPT can now read your entire iMessage history and send
arXiv:2608.18770v1 Announce Type: cross Abstract: Learning a reward model from human feedback and optimizing a policy against it is one approach to aligning AI systems with individual users. From a fa
arXiv:2608.18177v1 Announce Type: new Abstract: Continual learning has traditionally treated forgetting as a failure, emphasizing preservation of previously acquired knowledge as environments evolve.
arXiv:2608.18404v1 Announce Type: cross Abstract: We answer this question with Vector-Symbolic Policy Gradient (VSPG), a discrete-action actor that represents each action by a unit-norm hypervector an
arXiv:2608.18671v1 Announce Type: new Abstract: Egocentric video captures activities from the wearer's perspective, providing a direct view of human attention, hand--object interaction, and goal-direc
arXiv:2505.24313v3 Announce Type: replace Abstract: Weak-to-strong generalization (W2SG) is the phenomenon in which a powerful student model, trained on labels produced by a weaker teacher, ultimately
Today on Decoder, I’m talking with Robert Hart, The Verge’s London-based AI reporter, about what AI is doing to the field of mathematics and the existential crisis many lead mathematicians are having
we're launching new Fable safeguards for enterprises that work on your infrastructure so you have control over where your data lives & who has access to it we've been developing it alongside ~100 comp
We've been working on this with customers for a while. Mythos-class models require additional safety measures and enterprises need to meet their own privacy and compliance rules. Customers can own and
arXiv:2608.18397v1 Announce Type: new Abstract: Wearable stress classifiers can achieve strong average performance while failing completely for a particular individual. On WESAD, a Random Forest reach
arXiv:2608.18628v1 Announce Type: cross Abstract: Aligned vision-language models (VLMs) are designed to balance grounded visual reasoning with safe generation behavior. However, we observe a striking
arXiv:2608.18116v1 Announce Type: new Abstract: Vision-language models enable zero-shot classification through natural language prompts, but performance is sensitive to prompt formulation, especially
arXiv:2510.14000v2 Announce Type: replace Abstract: The growing demand for parking has increased the need for automated parking planning methods that can operate reliably in confined spaces. In restri
arXiv:2608.16893v1 Announce Type: cross Abstract: Expert surveys are widely used in security research to study practitioner workows and decision-making, yet recruiting domain experts - especially in S
arXiv:2606.12499v2 Announce Type: replace Abstract: We present AEM, an Action-Effect Memory pretraining framework for robot manipulation that learns compact temporal representations from vision-action
arXiv:2608.17929v1 Announce Type: new Abstract: Robust Markov decision processes optimize one policy against a set of plausible transition functions. This can be conservative when the unknown dynamics
arXiv:2608.16902v1 Announce Type: cross Abstract: The increasing integration of machine learning in healthcare has highlighted critical challenges related to fairness, transparency, and health equity.
arXiv:2602.20134v2 Announce Type: replace-cross Abstract: Epidemiological models increasingly rely on crowdsourced, self-reported behavioral data such as vaccination status, mask usage, and social dis
arXiv:2608.17987v1 Announce Type: cross Abstract: The rapid growth of social media has greatly influenced political discourse, highlighting the need to understand individual political ideologies and t
arXiv:2509.06777v2 Announce Type: replace Abstract: Graph Neural Networks (GNNs) suffer from oversquashing, where structural bottlenecks limit message propagation between distant nodes, hindering task
arXiv:2503.20182v2 Announce Type: replace-cross Abstract: As large language models (LLMs) increasingly function as human-like assistants exhibiting human-like personality traits, understanding their b
arXiv:2608.17638v1 Announce Type: new Abstract: What a reasoning model writes is only a partial record of the process that produces it. We introduce a two-level internal readout for mixture-of-experts
arXiv:2512.00074v4 Announce Type: replace-cross Abstract: Despite strong results on recognition and segmentation, current 3D visual pre-training methods often underperform on robotic manipulation. We
arXiv:2608.17496v1 Announce Type: new Abstract: Vision-language-action policies generalize broadly but provide no execution-time guarantees; classical model-based planners respect kinematic and geomet
arXiv:2608.18017v1 Announce Type: new Abstract: Improving flight safety with flight data requires not only accurate detection of risk events, but more importantly, clear interpretation of their underl
arXiv:2608.17070v1 Announce Type: new Abstract: With the growing deployment of machine learning models, formal guarantees of the robustness and fairness of these models have become increasingly import
arXiv:2608.16897v1 Announce Type: cross Abstract: Large-scale urban simulation plays a pivotal role in social science, traffic safety, and transportation policy. Recent work has shown that large langu
arXiv:2608.17234v1 Announce Type: cross Abstract: Multimodal large language models (MLLMs) are increasingly used to interact with screenshots, scanned documents, diagrams, and other visually grounded
Counterpoint: “If alignment issues are becoming big enough that OpenAI is willing to commit 20% of research inference compute to chain-of-thought monitoring [suggesting] that suggests that alignment i
ThreatDown, the business security arm of Malwarebytes Inc., said in new research published today that Kriminal, one of the newest and most popular tools in the criminal artificial intelligence market,
arXiv:2608.17050v1 Announce Type: cross Abstract: Methods for improving knowledge use in large language models typically fall into two regimes. Non-parametric retrieval offers flexible access to exter
arXiv:2608.17445v1 Announce Type: cross Abstract: Most large language model services use stateless defenses, which judge only the current request, to refuse harmful tasks. Decomposition attacks exploi
arXiv:2608.17657v1 Announce Type: new Abstract: Vision Transformers (ViTs) achieve state-of-the-art performance but carry massive computational overhead that restricts edge deployment. Although struct
arXiv:2608.17434v1 Announce Type: new Abstract: We study Gaussian regression over the explicit vector-valued Parhi--Nowak deep-RBV^2 architecture with depth L, width w, layer-sum variation budget A, a
arXiv:2608.17707v1 Announce Type: new Abstract: Audio-driven avatar generation requires realistic lip-sync, expressive motion, and real-time streaming. Recent work achieves the latter via self-forcing
arXiv:2608.17512v1 Announce Type: new Abstract: Although Large Vision-Language Models (VLMs) have significantly advanced embodied navigation, their direct deployment remains challenging, as existing m
arXiv:2603.02876v2 Announce Type: replace Abstract: Large Language Model personas, explicit profiles specifying a user's attributes, preferences, and behavioural tendencies, are increasingly used to s
arXiv:2608.17726v1 Announce Type: new Abstract: Bridge structures are regularly inspected for structural damage such as cracks and corrosion in order to ensure public safety and reduce maintenance cos
arXiv:2607.16660v2 Announce Type: replace-cross Abstract: The increasing adoption of Large Language Models (LLMs) as AI components in modern software systems introduces distinct security risks to the
arXiv:2608.17454v1 Announce Type: new Abstract: This paper presents a pipeline for analyzing media bias and framing in online news. The pipeline groups articles into topics and events, adds named-enti
arXiv:2608.17411v1 Announce Type: new Abstract: Group Relative Policy Optimization (GRPO) has become a widely used approach for post-training Large Language Models (LLMs) for reasoning. In GRPO, the g
A forum poll posed the question, “If OpenAI falls apart or is acquired at a big discount, will that on net help or hurt Anthropic?” Participants chose among three options—help, hurt, or neutral—reflec
arXiv:2608.14308v1 Announce Type: cross Abstract: We study score-matching diffusion models with a convolutional architecture. We argue that the inductive bias of locality means that the machinery of e
arXiv:2608.17051v1 Announce Type: cross Abstract: Secondary use of electronic health records requires de-identification, yet existing systems miss institutionally situated protected health information
It is genuinely painful to me that almost nobody — not even doomers who should care deeply — seems to be paying attention to this critical point that might have truly profound consequences for humanit
arXiv:2608.17994v1 Announce Type: new Abstract: Using LLMs as judges has become standard practice for evaluating model outputs at scale. This is particularly common for subjective, open-ended tasks su
arXiv:2602.01115v3 Announce Type: replace-cross Abstract: Diffusion-based visuomotor policies excel at modeling action distributions but are inference-inefficient, since recursively denoising from noi
arXiv:2608.17030v1 Announce Type: new Abstract: The massive overactuation in the human musculoskeletal system makes it challenging to train musculoskeletal models to generate human-like motion via rei
arXiv:2603.24126v2 Announce Type: replace Abstract: When language models are trained by reinforcement learning (RL) to write probabilistic programs, they can artificially inflate their marginal-likeli