Industry Adoption C

Memory benchmarks for LLM agents largely assume single-user settings, leaving shared-memory governance untested. GateMem benchmarks memory governance, such as access control and management, in multi-principal shared-memory agents.

Forecasting benchmarks for general-purpose AI usually inherit real-world events, making evaluation hard to control. ForecastBench-Sim introduces a simulated-world forecasting benchmark, enabling controlled assessment of AI forecasting ability.

Hitachi is expanding its partnership with OpenAI, pairing the code-analysis AI "Codex" with its own systems-development expertise. It aims to establish an AI-driven workflow that visualizes upstream specifications from existing code through migration testing, and also cites cybersecurity defense as a use case.

Preferred Networks continues developing its large language model PLaMo and shares how to use the latest PLaMo-3.0-Prime-β in real development work. Beyond training large models, it covers the many surrounding tasks involved in building high-performance LLMs in practice.

Cohere details how it ensures every tenant gets a fair share of compute in LLM serving, tackling the 'noisy neighbour' problem where one user monopolizes resources. The design allocates capacity fairly across tenants to deliver stable, predictable multi-tenant performance.

At self-service gas stations in Japan, staff still manually approve refueling after a safety check. Cosmo Oil Marketing and ELEMENTS have jointly developed a monitoring system in which AI judges whether to permit refueling, aiming to support that task. The article cites labor shortages and a declining number of service stations as background. Details are per the article and the companies.

The paper proposes VERITAS, a generator-verifier framework pairing a pre-trained generalist robot policy with a gradient-free visual verifier that evaluates actions at inference time, improving performance without extra training and enabling self-improvement.

The paper presents a methodology to systematically digitize and encode the legacy print Al-Mawrid Arabic-English dictionary using the ISO Language Markup Framework and TEI Lex-0, addressing a gap in Arabic lexical infrastructure by producing a standardized computational lexicon.

The paper examines test code generated by AI coding agents in open-source pull requests, arguing that test files lacking explicit assertions verify no behavior, so presence-based quality gates overestimate verification strength.

A robot's flash endurance is a non-renewable stock: each persisted write spends one of a few thousand program/erase cycles and never refills. The paper frames flash endurance as a wasting asset, proposes pricing it for embodied agents, and examines the limits of doing so.

The paper proposes a meta-knowledge reutilization framework for meta-reinforcement learning that learns task-level knowledge on a dynamics-simplified agent and transfers it to heterogeneous agents, using a Bayesian non-parametric prior to organize latent task modes.

Ternary Mamba applies grouped quantization-aware training to Mamba state space models with ternary (W1.58) weights and 16-bit activations, targeting efficient low-bit training and inference of sequence models while preserving accuracy.

As LLM-based agents enter autonomous scientific research, resisting pseudoscience matters. PseudoBench is an adversarial benchmark measuring how such agents may rapidly generate plausible yet misleading studies that contaminate academic literature.

INI-VPINN is a weak-form physics-informed neural network that naturally incorporates Neumann boundary and interface conditions into a variational formulation, targeting multi-material domains with geometric singularities.

E-commerce platforms struggle to understand customer behavior and predict future purchases. The study proposes predictive analytics using a hybrid Ret-DNN combined with an XGBoost model to forecast customer behavior.

Long chain-of-thought reasoning helps, but models often keep generating unnecessary reasoning after reaching a correct answer. Framing this as overthinking in GRPO-style RL post-training, the paper proposes dynamic rollout editing to reduce it.

SoftBank Group announced "Patching as a Service" on June 16, a cybersecurity offering built on OpenAI technologies such as "GPT-5.5 Cyber." It simulates attacks on corporate systems to find vulnerabilities, then proposes remediation plans and implementation end-to-end. SoftBank says it will prioritize select firms supporting Japan's critical infrastructure, while chairman Masayoshi Son stressed the gravity of the cyber threat.

EnvRL is a method that learns from environment dynamics in agentic reinforcement learning, leveraging the structure of agent-environment interaction to improve learning efficiency and performance.

Evaluating LLMs is important but can be fragile to small prompt changes. The paper proposes using prompt perturbation to achieve more reliable LLM evaluation over comparison graphs.

OpenAI introduced Deployment Simulation, a method to predict an AI model's behavior before deployment by using real conversation data to simulate responses, aiming to improve safety and evaluation accuracy. The claims are OpenAI's own and not independently verified.

Two companies including Osaka Gas sharply reduced the burden of hunting for documents by combining Notion with AI. Achieving 2,000 hours of monthly savings, the case turns buried information into organizational knowledge assets and highlights how to build systems that prevent over-reliance on individuals.

Generative AI is expanding beyond text, images, and video into 3D CAD, with environments emerging that draft 3D models from natural-language prompts alone. The article tries Autodesk Fusion's Autodesk Assistant to model a plastic bottle, illustrating both the promise and current limits of pairing generative AI with 3D CAD.

Sumitomo Mitsui Financial Group grew its Olive and Trunk services and unveiled a 50-billion-yen generative-AI investment plan. Once a bank lagging rivals in mobile a decade ago, it became an organization that repeatedly ships new ventures, finding the essence of new business in knowing when to exit.

Researchers built a 'value axis' for Qwen3-8B that captures whether its current strategy is likely to reach its goal. The axis separates high- and low-confidence rollouts, backtracking, and correct vs. corrupted code; steering it up suppresses self-correction while steering down induces exploration. DPO can raise the internal value of rewarded behaviors.

ROVE is an RL framework for post-training humanoid Vision-Language-Action models from imperfect human interventions. It pairs a human-in-the-loop data pipeline with Optimistic Value Estimation to prioritize high-value behaviors in mixed-quality trajectories, and adds cross-embodiment human videos to robustify value estimation.

Theory of mind (ToM), ascribing mental states to others for prediction and inference, is widely assumed essential for human-machine integration. Existing AI-ToM models address how to mentalize but leave when largely unaddressed. The paper asks under what situational and agent-level conditions ToM engagement is causally warranted in conflict, presenting a structural causal model as a directed acyclic graph that treats ToM as a mechanism activated by conditions rather than an always-on capacity.

Rovers rely on perception to maintain spatial maps encoding objects and sensor quality (range reliability, lighting artifacts, data density) to guide fusion, embedding updates, and navigation under partial observability. The paper presents CrossMaps, a real-time confidence-aware open-vocabulary semantic mapping pipeline that builds language-queryable maps from RGB-D data, extending VLMaps-style approaches with multi-scale CLIP embeddings, confidence-aware fusion, and a dual-memory architecture.

This paper studies reasoning with a Code Interpreter (CI) in LLMs from two angles: extrinsic properties (crucial tokens) and intrinsic properties (code-specific cognitive behaviors). It reports that stronger CI reasoning models show more crucial tokens and behaviors—especially verification, backtracking, and backward chaining—and explores leveraging these at inference and training time. Summarized neutrally from the abstract.

This paper reflects on a project-based master's course at the University of Bremen on engineering AI-enabled systems. It argues that machine learning courses emphasize model development while students lack experience in architectural design, deployment, and monitoring, and reports on the course's design and implementation.

An arXiv paper systematically studies tree linearization orders in Transformer Grammars, exploring breadth-first and a novel Production-Rule Traversal alongside the conventional depth-first approach. Summarized neutrally from the abstract.