Language models are increasingly deployed in interactive settings where users reason about facts over time rather than in isolation.In such scenarios, correct behavior requires models to maintain and update implicit temporal assumptions established earlier in a conversation.We study this challenge through the lens of temporal scope stability: the ability to preserve, override, or transfer time-scoped factual context across dialogue turns.We introduce ChronoScope, a large-scale diagnostic benchmark designed to isolate temporal scope behavior in controlled multi-turn interactions, comprising over one million deterministically generated question chains grounded in Wikidata.ChronoScope evaluates whether models can correctly retain inferred temporal scope when follow-up questions omit explicit time references, spanning implicit carryover, explicit scope switching, cross-entity transfer, and longer temporal trajectories.Through extensive evaluation of state-of-the-art language models, we find that temporal scope stability is frequently violated in controlled multiturn settings, with models often drifting toward present-day assumptions despite correct underlying knowledge.These failures intensify with interaction length and persist even under oracle context conditions, revealing a gap between single-turn factual accuracy and coherent temporal reasoning under sequential interaction.We make our dataset and evaluation suite publicly available at https://github. com/yashkumaratri/ChronoScope.

Effective content moderation systems require explicit classification criteria, yet online communities like subreddits often operate with diverse, implicit standards.This work introduces a novel approach to identify and extract these implicit criteria from historical moderation data using an interpretable architecture.We represent moderation criteria as score tables of lexical expressions associated with content removal, enabling systematic comparison across different communities.Our experiments demonstrate that these extracted lexical patterns effectively replicate the performance of neural moderation models while providing transparent insights into decision-making processes.The resulting criteria matrix reveals significant variations in how seemingly shared norms are actually enforced, uncovering previously undocumented moderation patterns including community-specific tolerances for language, features for topical restrictions, and underlying subcategories of the toxic speech classification.latex 1

Summary Los Alamos National Laboratory has deployed (over the last year and a half) a pair of Cray Shasta machines—a development testbed named Guaje and and production machine named Chicoma, which will soon comprise the bulk of LANL's open science research computing portfolio. In the process, we have encountered a number of problems and challenges in several realms—authentication and authorization, cluster health management, image management, and configuration management. Both independently and in collaboration with Cray/HPE, we have found solutions and brought the system into stable production. The presentation will discuss the solutions and how they came about, and issues we are working to resolve in the near future.

As a nonmuscular communication and control system for people with severe motor disabilities, brain–computer interface (BCI) has found several applications. Although a few empirical studies of BCI user performance do exist, little to no research has specifically evaluated the impact of contributing factors on user performance in the BCI applications. To that end, our within-subjects design compared the impact of two different types of interface (ABC interface vs. frequency-based interface) and three levels of screen size (computer monitor, global positioning system, and cell phone screen) of a P300-based BCI application, P300 Speller, on user performance (accuracy, information transfer rate, amplitude, and latency) and usage preference. Ten participants with neuromuscular disabilities such as amyotrophic lateral sclerosis and cerebral palsy and 10 nondisabled participants were asked to type six, 10-character phrases in the P300 Speller. The overall accuracy was 79.7% for the nondisabled participants and 28.7% for participants with motor disabilities. The results showed that interface type and screen size have significant effects on user performance and usage preference, with varying degree of impact to participants with and without motor disabilities. Specifically, participants typed significantly more accurately in frequency-based interface and computer monitor screen. The results of this study should provide invaluable insights to the future research of P300-based BCI applications.

Despite recent advances in brain-computer interface (BCI) development, system usability still remains a large oversight. The goal of this study was to investigate the usability of a P300-based BCI system, P300 Speller, by assessing how background noise and interface color contrast affect user performance and BCI usage preference. Fifteen able-bodied participants underwent a 2 (low and high interface color contrast) × 3 (low, medium, and high background noise level) within-subjects design experiment, in which participants were asked to type six 10-character phrases in the P300 Speller paradigm. The overall accuracy in the study was 80.2%. Participants showed higher accuracy, higher information transfer rate, bigger amplitude, and smaller latency in the high interface color contrast condition than in the low contrast condition. Participants had better performance in the noisy condition than in the quiet condition, but the background noise effects were not statistically significant in the present study. These results should give some insight to the real-world applicability of the current P300 Speller as a nonmuscular communication system, especially for individuals with severe neuromuscular disabilities.