Industry Adoption C
Showing 31–60 of 100
-
AI systems and the reproduction of (standard) language ideologies in World Englishes
-
Echoverse: Deep, evolving environments for computer-use agentsMicrosoft's Echoverse trains computer-use agents in evolving environmentsMicrosoft Research unveiled Echoverse, a set of deep, evolving environments for training computer-use agents that struggle with multi-step workflows such as email and customer support. Training in realistic settings aims to improve agents' ability to complete complex tasks.
-
EvoLib: Turning experience into evolving knowledgeMicrosoft's EvoLib turns experience into evolving knowledgeMicrosoft Research presented EvoLib, arguing that LLMs do not get smarter simply by remembering more. EvoLib converts task experience into reusable, evolving skills, aiming to let models continuously improve rather than just accumulate raw memory.
-
QuantWAMs: Calibrating at the Right Granularity for World Action Models
-
How Benchmarks Mis-Score Computer-Use Agents
-
ShadowDancer: Teaching Video World Models Any Action by Learning Unified Dynamics Representations from a Video and Its Shadow
-
Correlation between prosody and pragmatics: A case study of the discourse marker hālā `now' in Persian
-
Investigating three real-world incidents in our cybersecurity evaluationsAnthropic's Frontier Red Team probes three cybersecurity-eval incidentsAnthropic's Frontier Red Team published a review of three real-world incidents tied to its cybersecurity evaluations. The investigation examines potential misuse and the validity of its evaluation methods to strengthen the safety of frontier models.
-
Crusoe partners with Aalo on nuclear-powered AI data center deployment at Idaho National LabCrusoe teams with Aalo on SMR-powered AI data center at Idaho National LabCrusoe is partnering with Aalo to deploy a nuclear-powered AI data center at Idaho National Lab, DatacenterDynamics reported. Crusoe will install a Spark unit at the site, which is to be powered by a small modular reactor (SMR), reflecting growing interest in nuclear for AI compute.
-
Fully Inductive Cardinality Estimation
-
MonoVoc: Decoupling Geometry and Semantics for Lightweight Monocular Open-Vocabulary 3D Gaussians
-
MORFES: A Benchmark for Productive Inflectional Competence in Modern Greek
-
AI and Authenticity in Islamic Research: A Critical Evaluation of Generative AI Reliability, Hallucination, and Source Fidelity in Quranic, Hadith, and Fiqh Knowledge
-
Security of World-Model-Based Embodied AI: A Lifecycle of Threats, Defenses, and Evaluation
-
Integrating AI into Requirements Quality Learning in Software Engineering Education: A TPACK-Guided Empirical Study
-
ConMem: Contribution-Aware Memory for Long-Horizon Manufacturing Inspection Logs
-
Towards Practical Algorithm Selection for Unsupervised Domain Adaptation in Medical Imaging
-
Challenges in annotations by humans and LLMs: A case study of evaluative language
-
Echoverse: Deep, Evolving Environments for Training Computer-Use Agents at Scale
-
Stimulus-Evoked Network Dynamics in Human Cortical Organoids: From a Graph-Computational Framework to Repeated-Stimulation Depression
-
Building a User Foundation Model for the Open Web
-
TAPO: Transition-Aware Policy Optimization for LLM Agents
-
Advancing the price-performance frontier with GPT-5.6OpenAI advances the price-performance frontier with GPT-5.6OpenAI said GPT-5.6 pushes the price-performance frontier forward, pairing lower pricing with stronger capability. The update aims to make high-performance models more affordable and expand adoption across a wider range of use cases.
-
ChronoMem: Version Control and Semantic Rollback for Large Language Model Agent Memory
-
Harness-G: A Graph-Structured Harness for Search Agents
-
営業製作所、図面管理システム「ジーエン図面」の販路拡大へSB C&Sと契約Eigyo Seisakusho signs SB C&S to distribute its Gien Zumen drawing systemJapanese firm Eigyo Seisakusho signed a distributor agreement with SB C&S for its 'Gien Zumen' drawing-management system. By leveraging SB C&S's nationwide B2B sales network, the two aim to widen the product's distribution channels and grow the number of adopting companies.
-
Do You Really Need to Pretrain Q-Functions for Online RL Fine-Tuning?
-
Skillful forecasting of offshore winds from satellite scatterometer constellations
-
Investigating reservoir computing for branch predictionin pipelined processors using emerging CMOS memristor devices
-
Axe Compute secures $1.5bn compute contract with unnamed customerAxe Compute lands $1.5bn, five-year compute contract with unnamed customerDatacenterDynamics reports that Axe Compute has secured a $1.5 billion compute contract with an unnamed customer, with the agreement set to run for five years. The customer's identity, workload and facility details are outside the excerpt and unconfirmed.