Babak's Daily Briefing

Monday, June 29, 2026

Sources: 20 | Total Articles: 96

• 1.The Hidden 6G Bottleneck: RF Hardware Design Is Becoming a Strategic Race

  As 5G-Advanced matures and 6G research moves closer to implementation, the wireless industry faces a deeper challenge than spectrum, standards or AI-native network architecture. Future wireless systems will depend on whether the industry can design, validate and manufacture increasingly complex RF modules fast enough.

• 2.6G in Dalian: What the Latest 3GPP Meetings Reveal About the Future Radio and Network

  The 6G physical layer is starting to converge. The protocol stack is being simplified in meaningful places. But the most consequential architecture decisions are now moving toward the June plenary in Singapore.

• 3.RF Digital Twins: Why 5G-Advanced and 6G Need Predictive Simulation

  RF Digital Twins: Why 5G-Advanced and 6G Need Predictive Simulation As wireless systems become more tightly coupled across…

• 4.Evaluating 6G PHY Evolution: What the Industry Is Really Trying to Solve

  Summary available at source link.

• 5.Amazon’s Globalstar deal gives Amazon Leo a faster path into D2D

  Amazon’s planned acquisition of Globalstar is about far more than satellites. It gives Amazon Leo a faster path into direct-to-device connectivity, combining spectrum, operational assets, and Apple-facing service continuity in a move that could reshape the hybrid terrestrial-NTN landscape.

• 1.SAFARI: Scaling Long Horizon Agentic Fault Attribution via Active Investigation

  As autonomous agents tackle increasingly complex multi-step, multi-agent tasks, their execution trajectories have scaled beyond the constraints of even the largest context windows. Current methods for effectively diagnosing agent failures load the full trajectory into an LLM's context window, which suffers from attention dilution and fails when agentic traces inevitably exceed context limits. To address this, we introduce SAFARI (Scaling long-horizon Agentic Fault AttRibution via active Investigation), a framework that replaces linear context loading with a tool-augmented diagnostic loop. By equipping LLMs with a specialized toolbox to read and search trajectory segments alongside a persistent Short-Term Memory (STM) for cross-turn reasoning, SAFARI effectively decouples diagnostic accuracy from architectural context limits. Our experimen...

• 2.Project Auto-World: Towards Automated Benchmarking of Neural Relational Reasoners

  Reasoning about relational structures remains a significant challenge for neural models, particularly when they must systematically apply learned knowledge to problem instances that are harder than those seen in training. Progress is hampered by the difficulty of evaluating such generalization, since a priori, it is rarely clear what makes an instance hard. We study how this issue can be addressed by using large language models (LLMs) to automate benchmark generation, learning to produce increasingly challenging instances in an end-to-end manner. Concretely, given a world parametrized by Datalog rules, and an Edge Transformer as the reasoning evaluator, we use LLM-driven evolutionary search (based on FunSearch) and autonomous agentic search to discover sampling functions that yield hard problem instances. We also show that the Edge Transf...

• 3.AgenticOS: An Intent-Oriented Secure Operating System Architecture for Autonomous AI Agents

  Traditional OS security models based on "resource exposure plus permission checks" face structural challenges as LLM-driven autonomous agents acquire capabilities for planning, tool use, network access, and code execution. Once an agent runtime is compromised through prompt injection or malicious tool outputs, an attacker can compose POSIX-style resource primitives into behaviors far beyond the user's task authorization. To address this, we propose AgenticOS, an intent-oriented secure OS architecture that consolidates delegable, auditable software capabilities into OS-native ones rather than replacing all applications. The core insight is to reframe the OS from a "resource manager" into an "intent filter": instead of requesting low-level resources directly, agents submit structured intent declarations, from which the system synthesizes a ...

• 4.ORAgentBench: Can LLM Agents Solve Challenging Operations Research Tasks End to End?

  Large language models are increasingly deployed as autonomous agents for multi-step tasks in executable environments, yet their ability to perform realistic operations research (OR) work remains unclear. Existing OR evaluations often decouple modeling from solving, rely on pre-formalized or text-only instances, and rarely test the full workflow from operational artifacts to validated decisions. In this work, we introduce ORAgentBench, an execution-grounded benchmark for evaluating autonomous agents on challenging end-to-end operations research tasks. It contains 107 human-reviewed tasks across diverse operational scenarios, each packaged in an isolated environment with a natural-language brief, multi-file data, configuration artifacts, and a required submission schema. Agents must write and run solution code, and their submissions are eva...

• 5.Deontic Policies for Runtime Governance of Agentic AI Systems

  Autonomous agentic AI systems driven by Large Language Models (LLMs) introduce a new class of security, privacy, and compliance challenges: an agent that can invoke tools, manipulate data, install software, and coordinate with peer agents across organizational boundaries must be constrained not just by authentication and access control, but by the full structure of enterprise governance. This includes specifying what agents are permitted and prohibited from doing, what they areobliged to do after certain actions (e.g., notify the CISO), under what conditions a standing obligation may be waived, and which rules take precedence when policies conflict. This governance problem exceeds what current policy engines provide. Systems such as XACML, Rego, and Cedar address only the permit/prohibit subset of this governance structure. They do not pr...

• 1.Formalizing Latent Thoughts: Four Axioms of Thought Representation in LLMs

  arXiv:2606.27378v1 Announce Type: new Abstract: We introduce an axiomatic evaluation framework for latent thought representations in LLMs, comprising metrics that are independent of downstream benchmark scores and reveal representational failures that benchmark accuracy masks. Existing evaluations conflate representation quality with model capacity. Therefore, failures cannot be attributed to the representation rather than to the model that processes it. We formalize four functional axioms (Causality, Minimality, Separability, and Stability) and define a quantitative measure for each, computed directly on the representation independently of downstream accuracy. We audit open-weight LLMs across 23 reasoning tasks (e.g., Spatial Reasoning, Factual QA). We find that no candidate satisfies all four axioms simultaneously, that the representat...

• 2.Position: The Term "Machine Unlearning" Is Overused in LLMs

  arXiv:2606.27379v1 Announce Type: new Abstract: Large language models increasingly face demands to "forget" training data, knowledge, or behaviors due to regulatory deletion obligations, copyright/licensing disputes, and safety or product-policy requirements. This position paper argues that machine unlearning is overused as a term in LLM research and should be reserved for dataset-defined deletion: removing the training influence of a precisely specified forget set such that the resulting model is approximately indistinguishable from retraining without that data. We contend that many tasks currently labeled "unlearning" (e.g., refusal for harmful requests, entity/knowledge removal, or targeted suppression) pursue different, often policy-dependent objectives and therefore require different terminology and baselines (e.g., alignment, suppr...

• 3.A Survey of Automated Presentation Coaching: Systems, Methods, and Open Challenges

  arXiv:2606.27380v1 Announce Type: new Abstract: Automated coaching for oral presentations sits at the intersection of computer-assisted pronunciation training (CAPT), prosody modeling, and speech synthesis, yet no prior work has systematically surveyed and compared existing systems along these dimensions. This survey reviews and categorizes automated presentation coaching systems, spanning pronunciation tutors, fluency and prosody coaches, multimodal trainers, and conference Q&A practice tools. We introduce a five-dimensional task taxonomy - covering segmental pronunciation, lexical stress, suprasegmental prosody, pacing, and content faithfulness - and explicitly map surveyed systems onto it to reveal coverage gaps. We further review the core technical methods these systems employ: TTS-based exemplar generation and diagnostic methods for...

• 4.Causal Connections: Leveraging Multilingual Fine-Tuning for Financial QA@FinCausal 2026

  arXiv:2606.27446v1 Announce Type: new Abstract: This paper describes team HSA_CORAL's submission to the FinCausal 2026 shared task on extracting cause-effect relations from financial narratives via extractive question answering in English and Spanish. We compare three modeling families: (i) encoder-only token tagging with multilingual BERT, (ii) encoder-decoder generation with multilingual BART, and (iii) decoder-only LLMs (Llama 3.1 and GPT variants) using prompt refinement, few-shot demonstrations, and supervised fine-tuning. Across settings, prompting and few-shot examples yield competitive performance, while supervised fine-tuning provides the largest gains. Our best system, GPT-4.1 Mini fine-tuned on combined English and Spanish training data, achieves a tied highest score on the English subtask (score 4.8140) and ranks third on Spa...

• 5.Developmental approach reveals the statistical learning of Neural Language Models: Transformers generalize from the most abstract statistical patterns

  arXiv:2606.27460v1 Announce Type: new Abstract: In this study, we use a developmental approach to investigate the statistical learning and mental representation of neural language models (NLM). A series of Generative Transformer models are trained on a synthetic grammar. The model states are saved at multiple stages in the course of training. Through analyzing how the internal representations of these models change in the developmental path, we found that NLMs acquire the most abstract global statistical knowledge at the beginning of learning and later acquire the relatively local statistical dependencies. This learning path contains many over-generalizations from the very beginning and these over-generalizations are gradually constrained in the later stage of learning. Based on this observation, we propose a new framework to explain the...

• 1.OverFlowLight: Real-Time Gridlock Prevention and Traffic Signal Optimization for Urban Intersections

  arXiv:2606.27381v1 Announce Type: new Abstract: Queue overflow, a severe consequence of urban traffic congestion, occurs when vehicle queues exceed intersection capacity, obstructing upstream traffic and triggering cascading gridlocks. Prevailing traffic signal control (TSC) algorithms, primarily optimized for throughput, often fail to address overflow during peak hours, exacerbating congestion and creating safety hazards. We propose OverFlowLight, a real-time framework designed to preemptively resolve overflow and enhance overall TSC performance. It first introduces a mechanism to accurately detect overflow in real-time by leveraging multi-modal sensing from cameras and radars. Upon detection, it dynamically generates and inserts dedicated overflow phases into the signal cycle to clear the blocking queues. This is orchestrated by a hybri...

• 2.RANSAC Scoring Done Right

  arXiv:2606.27385v1 Announce Type: new Abstract: The most widely used RANSAC variants score candidate models by counting inliers or summing per-point scores that saturate beyond a residual threshold. Every such score requires a user-supplied parameter that is a function of the inlier scale, which must itself be estimated from contaminated data. We remove this dependence by reversing the usual order of inference: rather than estimating the scale and then scoring against it, we marginalize the inlier scale analytically in closed form under a conjugate Inverse-Gamma prior for a fixed inlier partition, then optimize over partitions. A single closed-form expression spans the non-informative Jeffreys limit and informative empirical-Bayes priors, so the same score adapts across data-rich and data-scarce regimes without any change to the algorithm...

• 3.Unified Zero-Shot Time Series Forecasting: A Darts Foundation

  arXiv:2606.27438v1 Announce Type: new Abstract: Since its initial release in 2020, Darts has become a widely used open-source Python library for time series analysis. A series of foundation models have recently claimed accuracy improvements in zero-shot forecasting, promising a paradigm shift from training custom models to harnessing pre-trained general-purpose forecasters. Foundation models, however, are often released as isolated packages with fragmented interfaces and limited interoperability with common tooling, making joint evaluation and integration within complete pipelines difficult. In Darts, we developed a unified $\texttt{FoundationModel}$ class collection (Chronos-2, TimesFM 2.5, TiRex, PatchTST-FM) that provides standardized, full-cycle forecasting interfaces with minimal external dependencies for integrating foundation model...

• 4.PairSAE: Mechanistic Interpretability from Pair Representations in Protein Co-Folding

  arXiv:2606.27440v1 Announce Type: new Abstract: Foundation models for structural biology have achieved remarkable performance in predicting biomolecular structure and show promise for the design of proteins and small molecules. Yet understanding which internal features drive their outputs remains challenging. Standard sparse autoencoders (SAEs), effective on transformer-style sequence embeddings, do not transfer cleanly to pairformer-like architectures: naively operating on pairwise representations yields a quadratic blow-up of features and obscures concepts distributed jointly across sequence and pair representations. We introduce PairSAE, which summarizes pairwise tensors via an N-mode SVD into token-wise interaction roles, then uses a sparse autoencoder to learn a shared set of token-level features that decode into both sequence and pa...

• 5.Learning in Markovian bandits with non-observable states and constrained decision epochs

  arXiv:2606.27448v1 Announce Type: new Abstract: This paper studies the problem of regret minimization in Markovian bandits with \emph{non-observable states} and possibly \emph{constrained} decision epochs. The focus is restricted to a ``pure'' regret benchmark, that compares the performance of the learning algorithm to the best \emph{pure policy} which -- akin to optimal policies of stochastic bandits -- picks the optimal arm from start to finish without ever switching. We introduce a generalization of rested Markovian bandits, \emph{self-degrading Markovian bandits}, for which pure policies are always asymptotically optimal.We show that without prior knowledge on the underlying bandit, the regret of algorithms that switch arms rarely necessarily scales super-logarithmically for every bandit, i.e., as $\omega(\log(T))$, where $T$ is the l...

• 1.Support-Constrained RL Enables Real-World Policy Improvement without Real-World Experience

  arXiv:2606.27475v1 Announce Type: new Abstract: Robots trained on real world data tend to be imprecise, slow, and brittle to perturbations. Improving these policies with reinforcement learning (RL) is an appealing alternative, but this process often requires expensive training in the real world. Performing policy improvement in simulation instead provides a far cheaper alternative, but unconstrained RL in simulation can exploit contact and dynamics mismatches, resulting in unsafe behaviors that do not transfer to hardware. Common forms of regularization can furthermore limit improvement by overconstraining to an imperfect behavior prior. In this work, we propose Support-Constrained Off-Domain REinforcement (SCORE), a real-to-sim-to-real framework that constrains RL in simulation to the support of a generative policy pretrained on real dat...

• 2.AO-ARC: Almost-Surely Asymptotically Optimal Multi-Robot Motion Planning with ARC

  arXiv:2606.27495v1 Announce Type: new Abstract: We present AO-ARC, an anytime multi-robot motion planning (MRMP) method that achieves initial solution times on par with state-of-the-art MRMP feasibility solvers while converging faster and more reliably than existing anytime MRMP methods as the number of robots increases. AO-ARC adapts the AO-x meta-algorithm for converting feasibility solvers into anytime algorithms by iteratively calling the original ARC method on bounded MRMP instances under a makespan cost metric. This exploits the adaptive (de)coupling of ARC while maintaining the consistent cost bound across robot (de)compositions needed for AO-x. We provide theoretical analysis proving the asymptotic optimality properties of AO- ARC and conduct empirical evaluation on a set of 2D scenarios with different levels of coordination compl...

• 3.Spacecraft Fiducial Marker for Autonomous Rendezvous, Proximity Operations, and Docking

  arXiv:2606.27566v1 Announce Type: new Abstract: Robotic operations in space are challenging due to the harsh environment and the high cost of failure. Fiducial markers provide visual references that aid autonomous rendezvous, proximity operations, and docking for space robots. However, existing fiducial markers are mostly single-scale and largely designed for terrestrial robotics. Such markers leave the camera's field of view at close range, precisely during the proximity and docking phases where reliable tracking is most critical. This paper presents AstraTag, a fiducial marker designed for autonomous on-orbit robotic operations. The marker template is based on a square Spidron pattern whose recursive, self-similar structure enables detection across multiple spatial scales. Marker identification uses a 48-bit signature derived from trian...

• 4.SceneBot: Contact-Prompted General Humanoid Whole Body Tracking with Scene-Interaction

  arXiv:2606.27581v1 Announce Type: new Abstract: Current humanoid reinforcement-learning policies excel at free-space motions but struggle with contact-rich tasks, as pure kinematic tracking cannot resolve the physical ambiguities of interacting with objects and uneven terrain. To address this, we introduce SceneBot, a unified motion-tracking framework capable of handling freespace locomotion, terrain traversal, and whole-body manipulation. SceneBot conditions a single policy on both reference motions and per-link contact labels, explicitly defining expected environmental interactions. To overcome the lack of annotated interaction data, we propose a hindsight scene reconstruction approach that infers scene-interaction graphs from retargeted human motion. Trained on 7.5 hours of this reconstructed, contact-rich data, SceneBot successfully g...

• 5.Learning to Throw: Agile and Accurate Cable-Suspended Payload Delivery with a Quadrotor

  arXiv:2606.27603v1 Announce Type: new Abstract: Quadrotors offer the agility needed to rapidly transport suspended payloads during time-critical applications, including search-and-rescue and medical delivery. While suspended-payload transport and traversal for these missions are well studied, the highly dynamic targeted release of the payload remains comparatively underexplored. State-of-the-art approaches typically rely on model-based trajectory optimization and tracking; however, these methods often yield sub-optimal performance due to conservative feasibility constraints, tracking errors, and the inherent difficulty of analytically modeling flexible rope dynamics. To overcome these limitations, we propose a hybrid simulation framework that couples a high-fidelity analytical quadrotor model with a physics solver for complex rope and pay...

• 1.The Decision Geometry of Covariance Estimation for the Global Minimum-Variance Portfolio under Heavy Tails

  The global minimum-variance portfolio (GMVP) is the canonical decision built from an estimated covariance matrix, yet covariance estimators are universally evaluated by matrix-norm loss, which is not the object the decision depends on. We characterise exactly how covariance-estimation error maps into GMVP suboptimality. We prove an exact regret identity and a non-asymptotic bound showing decision regret depends on the estimation error only through its action on the portfolio weights, scaled by portfolio concentration and the conditioning of the true covariance. From this we derive the decision geometry: GMVP regret is invariant to a (p-1)-dimensional projection of the p^2-dimensional error matrix, with invariance to the covariance-scale direction as an exact special case. We then apply the framework to heavy-tailed returns (tail index kap...

• 2.Data-Driven Duration Management -- Term Structure Forecasting Using Machine Learning

  This paper compares different methods for forecasting the term structure of U.S. and European zero-coupon government bonds using both traditional econometric and Machine Learning (ML) approaches. We compare classical models (e.g., Dynamic Nelson-Siegel (DNS) and Principal Component Analysis (PCA)) with different Neural Network (NN) architectures, including those inspired by the classical models, on the U.S. Treasury market and bonds issued by the European Central Bank (ECB). To enhance predictive performance, macroeconomic variables are incorporated. The findings for both markets are separately analyzed and compared. To this end, we propose a robust model evaluation framework combining statistical accuracy metrics - such as RMSE, MAE, and directional accuracy - with the economic relevance of a quantitative bond trading strategy. Results s...

• 3.OpenFinGym: A Verifiable Multi-Task Gym Environment for Evaluating Quant Agents

  Although large language model agents are increasingly applied to quantitative-finance workflows, their evaluation remains fragmented across isolated tasks, while the financial relevance of benchmark tasks is often overlooked. Yet financial workflows are inherently multi-stage, spanning interdependent tasks such as forecasting, strategy construction, risk management, and trading. Existing platforms typically focus on a single task, and can therefore overstate agent competence and fail to reveal weaknesses in generalization, real-market interaction, and financially meaningful decision-making. We introduce OpenFinGym, a unified gym environment for quantitative-finance agent development that covers forecasting, market generation, real-time trading, and fraud detection under a single execution and verification interface. OpenFinGym additionall...

• 4.The Inference-Compute Frontier and a Latency-Efficient Architecture for Limit Order Book Prediction

  We study whether a scaling-law-style inference-compute frontier appears in limit order book prediction. Using FI-2010 and a suite of models ranging from small decision trees to neural LOB architectures, we find that the realized empirical frontier of predictive loss versus structural forward work is well summarized by a power law. In particular, with MLPLOB held out as an architecture family, a power-law fit to the low- and mid-compute non-MLPLOB frontier extrapolates across multiple orders of magnitude and attains $R^2=0.941$ on the excluded high-compute MLPLOB target frontier. A similar exercise in latency space gives substantially weaker results, showing that latency is not merely noisy compute. We use this gap to motivate FastBiNLOB, a dense axis-separable LOB mixer built from hardware-friendly temporal and feature mixing operations...

• 5.Hierarchical Graph Learning for Calendar Spread Strategies in Commodity Futures Markets

  Commodity futures can be represented hierarchically, with underlying assets at the upper level and individual futures contracts at the lower level. Entities at each level can be connected by edges reflecting inherent correlations, with cross-level edges capturing contract-to-underlying asset connections. Building on our observations of these structures, we propose a hierarchical graph learning approach for calendar spread (CS) strategies in commodity futures markets, addressing two significant gaps in the machine-learning literature: (i) the absence of learning-based methods for CS strategies in futures markets, and (ii) the lack of consideration of maturity-dependent interrelationships across commodity futures. We first establish the efficacy of CS strategies by analytically showing that CS strategies can possess higher risk-adjusted ret...

• 1.Telco Common Corpus: The largest open, verified data commons for telecom AI

  Summary available at source link.

• 2.GSMA Launches Global Satellite Regulatory Playbook to Help Policymakers Build Future-Ready Connectivity Frameworks

  Summary available at source link.

• 3.GSMA welcomes China Tower to advance AI-ready mobile infrastructure

  Summary available at source link.

• 4.Scaling Telco AI from Ambition to Reality

  Summary available at source link.

• 5.New GSMA Report: Digital Reforms Could Unlock FCFA 870 Billion and Connect Over 540,000 More People in the Republic of the Congo by 2030

  Summary available at source link.

• 1.LLawCo: Learning Laws of Cooperation for Modeling Embodied Multi-Agent Behavior

  Embodied agents operating in decentralized and partially observable environments have attracted growing attention in recent years. However, existing large language model (LLM)-based agents often exhibit behaviors that are misaligned with their partners or inconsistent with the environment state, leading to inefficient cooperation and poor task success. To address this challenge, we propose a novel framework, Learning Laws of Cooperation (LLawCo), that enables embodied agents to autonomously align with both their partners and task objectives. Our framework allows agents to reflect on past failures to extract misaligned behavioral patterns, which are used to derive high-level behavioral laws, such as "Talk when necessary" and "Wait for partner." These laws are explicitly incorporated into the agents' chains of thought via supervised fine-tu...

• 2.Tandem Reinforcement Learning with Verifiable Rewards

  Reinforcement learning with verifiable rewards (RLVR) has significantly improved the reasoning capability of large language models, reaching expert or even superhuman performance in domains such as competition math. However, whether weaker agents and humans can actually harness this capability is far less certain, with RLVR documented to drift reasoning toward idiosyncratic patterns such as poor readability and language mixing. Tandem training is a recently introduced paradigm that targets this compatibility problem: a trained, stronger senior co-generates each rollout with a frozen, weaker junior, and the two are rewarded as a team, so the senior is pushed to reason in ways the junior can follow. Yet this paradigm has so far been demonstrated only in proof-of-concept settings, leaving open whether it scales to the long chains of thought ...

• 3.AirGroundBench: Probing Spatial Intelligence in Multimodal Large Models under Heterogeneous Multi-View Embodied Collaboration

  In recent years, multimodal large language models (MLLMs) have shown strong potential for embodied intelligence, yet their ability to maintain geometrically consistent spatial understanding across heterogeneous views remains under-evaluated. Existing benchmarks largely focus on single-agent, single-view perception, leaving a gap in the systematic assessment of collaborative air-ground settings, where multi-scale observations are complementary but introduce scale mismatch, asymmetric occlusion, and reference-frame inconsistencies. We present AirGroundBench, a diagnostic benchmark for evaluating multi-view spatial intelligence in heterogeneous UAV-UGV collaboration. AirGroundBench is built from 11 high-fidelity simulated environments with 1,021 synchronized air-ground observation pairs, yielding approximately 62,000 dual-view, four-option s...

• 4.HumanMoveVQA: Can Video MLLMs reason about human movement in videos?

  Despite the rapid advance of Multimodal Large Language Models (MLLMs) in high-level video understanding, a fundamental bottleneck remains: these models collapse complex human motion into coarse semantic labels. Existing benchmarks mostly focus on scene-centric events or local joint articulations, failing to probe global human motion in space over time (trajectory and orientation changes). We introduce HumanMoveVQA, the first comprehensive benchmark designed to evaluate global trajectory and orientation reasoning from an exocentric perspective. Our benchmark utilizes a first-frame anchored world coordinate system, preserving translation and rotation relative to a fixed starting point. We propose a scalable, multi-stage pipeline that lifts 2D video observations into world-consistent 3D motion tracks to generate over 10K structured question-...

• 5.Understanding How MLLMs Describe Artworks Using Token Activation Maps

  Multimodal Large Language Models (MLLMs) describe artworks with remarkable fluency, yet the visual reasoning behind their outputs remains opaque. When an MLLM names a style, identifies a subject, or recognizes an iconographic symbol, does it ground each claim in the relevant region of the canvas, draw on an undifferentiated visual signal, or rely primarily on textual priors? We study this using the Token Activation Map (TAM), which produces, for each generated token, a heatmap isolating the visual evidence specific to that token from prior-context interference. Applying TAM to a curated set of paintings spanning multiple periods and genres, we analyze grounding patterns across five semantically distinct token categories: common visual objects, style descriptors, metadata, iconographic tokens, and affective expressions. We find that visual...

• 1.From Tokens to States: LLMs as a Special Case of World Models and the Continuous Path Beyond

  The AI community has framed the relationship between large language models (LLMs) and world models as a dichotomy: LLMs predict tokens; world models simulate reality. Yann LeCun argues in 2022 that reaching general intelligence requires abandoning autoregressive token prediction in favour of latent-space architectures. This framing is unnecessarily binary. Two claims will be defended. First, LLMs are a degenerate special case of world models: the state space is the set of all token sequences, the only action is appending one token, and world models are therefore a strict generalisation of LLMs, not a replacement. Second, there is a natural continuous spectrum from NTP to JEPA, with multi-token prediction, future-summary prediction, and next-latent prediction as intermediate stations already populated by current research. Moving along this...

• 2.Training Observable Control Policies to Expose Agent State Through Actions

  Physical or operational constraints often impose communications limitations on autonomous agents. Such limitations complicate monitoring or multiagent coordination. Even when strong communications are absent, some information may still be available. The remainder of the relevant agent state may be reconstructed via estimation. The actions taken by an agent are a potential source of information -- as the agent interacts with the environment, these actions may be observed even in the absence of explicit communication. We investigate using actions to estimate the state of an agent, using reinforcement learning to develop policies which make the estimation problem more tractable. Policy observability is encouraged through the training reward and is analyzed using simulation of the trained agent. In an aircraft tracking problem a policy with e...

• 3.SatSplatDiff: Geometry-preserving generative refinement for high-fidelity satellite Gaussian Splatting

  Gaussian Splatting has been recently explored for satellite 3D reconstruction, demonstrating flexibility and efficiency in representing radiometrically diverse satellite scenes. However, the limited top viewpoint of satellite imagery results in insufficient supervision on building facades, leaving surface holes and degraded visual fidelity. Generative refinement, which leverages pretrained generative priors to iteratively refine and update the rendered images used as supervision targets, has recently been investigated to improve the visual fidelity of Gaussian-rendered images. However, since these models refine each view independently, the resulting images can generate hallucinations and break photo-consistency, leading to geometric degradation. To address these limitations, we propose SatSplatDiff, which aims to minimize geometric degrad...

• 4.Reasoning Quality Emerges Early: Data Curation for Reasoning Models

  Supervised fine-tuning (SFT) on a small, high-quality set of long reasoning traces is an effective approach for eliciting strong reasoning capabilities in Large Language Models (LLMs). However, existing methods for curating high-quality SFT data rely heavily on strong reasoning models to filter examples based on diversity and difficulty, making the curation process costly while often yielding suboptimal data quality. In this work, we show that diverse and challenging reasoning examples can be identified using only the initial reasoning tokens. Specifically, we demonstrate that difficult problems can be reliably detected based on the loss of the first 100 reasoning tokens evaluated at a randomly perturbed checkpoint of the pretrained model. We further show that examples exhibiting similar loss patterns over their first 1k reasoning tokens ...

• 5.Anatomy-Guided Residual Motion Diffusion for Controllable 4D Cardiac MRI Synthesis

  Developing robust artificial intelligence models for 4D (3D + time) medical imaging is constrained by limited annotated data, inter-device domain shifts, and privacy restrictions. To address this, we propose a 4D controllable generative framework for anatomically consistent data augmentation. A semi-supervised variational autoencoder learns a compact latent representation of anatomical volumes while jointly predicting aligned segmentation masks in a unified framework. Anatomical structure is then disentangled from temporal dynamics through a cascaded latent diffusion model (LDM). A static LDM generates subject-specific anatomy conditioned on clinical priors (diagnosis and volumes measures) and a subsequent motion LDM estimates residual latent motions, ensuring strict temporal coherence across the 4D sequence. The proposed approach was eva...

• 1.ConlangCrafter Turns AI to Imagining Languages

  There are over 7,000 natural languages today, but that doesn’t stop people from occasionally making up completely new ones. These constructed languages, or conlangs , include Dothraki , Klingon , and various Elvish languages . Now, an AI model called ConlangCrafter is also capable of generating new languages—and it is particularly good at it. In a paper published 27 June in the Proceedings of the Association of Computer Linguists, researchers analyzed ConlangCrafter’s language generation abilities, reporting that it can develop a diverse array of novel languages that consistently abide by their rules. How ConlangCrafter Creates New Languages In previous work, Gašper Beguš , an associate professor of linguistics at the University of California, Berkeley, showed how large language models (LLMs) can analyze languages to the same extent as mo...

• 2.Why Does a Bank Need a Chief Scientist?

  This article is brought to you by Capital One . After five years leading natural language understanding and eventually the entire Alexa AI organization at Amazon, Prem Natarajan made a nontraditional move: He became Chief Scientist at a bank. Not just any bank: Capital One, a financial institution serving over 100 million customers, helping everyday Americans manage their financial lives. For Natarajan, a veteran of DARPA-funded research and academia who had watched machine learning evolve from task-specific applications to foundation models, the logic was clear. Some of the most interesting advances in AI research and deployment were shifting from big tech’s horizontal platforms to industry verticals like finance, where the most complex problems aren’t just building models but making AI work under the constraints of real-world customer p...

• 3.What it Means to Be a Mathematician When AI Does the Math

  In the mid-noughties, when music by the Killers and Franz Ferdinand blared out of every pub and nightclub I passed, I spent my days and nights struggling through a Ph.D. in applied mathematics . My research focused on simulating how special light waves interact in liquid crystals and using simple equations to approximate and understand those interactions. When I look back at my thesis now, liquid crystal technology is old hat, and I imagine my work could be completed with AI assistance in a matter of days—maybe hours. But the same cannot be said for the work of the pure mathematics Ph.D. students with whom I shared a cramped office at the University of Edinburgh. At the time, I felt sorry for these colleagues, who day after day sat at their desks, seemingly tearing their hair out and making no progress. (Though I was struggling too, I was...

• 4.AI Is Designing Radio Chips That Humans Couldn’t Even Imagine

  Summary RFIC design is a complex “ dark art ” that limits progress in wireless technologies like 5G, autonomous vehicles, and satellite communications. Princeton researchers use reinforcement learning and inverse design to rapidly create RFICs from scratch. Diffusion models rapidly generate novel or human-interpretable RF layouts, achieving record performance and drastically reducing design time. Future progress needs large, shared chip design datasets and open ecosystems so AI can learn universal electromagnetic and circuit behaviors. Take a moment and try to imagine your life without the wireless advances of the past three decades. Have you lost your luggage? What a shame AirTags have not been invented. The airline representative has promised to call with updates, so settle in for a long wait by the kitchen telephone, because there are ...

• 5.AI Is Learning to Read the Room

  Imagine sitting down at your desk and logging in for a performance review, with an AI system analyzing the conversation. You’ve been working long hours, balancing deadlines, and your manager asks how you’re doing. You say you’re fine, and maybe even smile, but there’s a hint of hesitation and your voice wavers. As you shift your posture, your shoulders slump. These are subtle cues that to the human eye might hint at underlying stress. But to an AI model that’s been trained only to categorize emotions as “happy” or “sad,” such nuances are likely lost. It logs the words and a smile and moves on—and unless your human manager intervenes, the fact that you’re tired, unfocused, and maybe a couple of days from burnout never enters the equation. “ Emotion AI ,” which estimates how people feel based on facial expressions, voice tone, and behavior,...

• 1.Transforming Investing With AI at Franklin Templeton

  Patrick George/Ikon Images What would you do with artificial intelligence if you were confident that it would transform your industry? What actions would you take if you felt that you were at an inflection point in that transformation? Would you try to be an early proponent of AI-first in your industry, or a fast follower? […]

• 2.Redefine What ‘Professionalism’ Means

  Matt Kenyon/Ikon Images “Professionalism” encompasses the broad set of shared beliefs and expectations about how people within an industry or workplace should interact with one another: Think communication style, punctuality, or meeting etiquette. But opinions differ: Cameras on? Cameras off? Do meetings start precisely on the hour? Is arriving a few minutes late acceptable or […]

• 3.Three Approaches to Measuring and Managing AI ROI

  Matt Harrison Clough/Ikon Images After several years of AI experiments and pilot initiatives, a crucial question remains open for most companies: How much of a return — and what kinds of returns — are we getting from all of this AI investment? To many executives, AI ROI still often feels more like art than science: […]

• 4.Resolving Muddled Objectives in Corporate Venture Capital

  Carolyn Geason-Beissel/MIT SMR | Getty Images The Research The authors compared the approaches of prominent corporate venture capital (CVC) units, including those owned by Intel, Cisco, General Electric, Siemens, NTT Docomo, Hitachi, Panasonic, and Sompo. They examined 59 of the most active CVCs tracked by research firm CB Insights from 2017 through 2024 and mapped […]

• 5.Leaders at All Levels: How DBS Bank Makes Everyone an Innovator

  DBS Bank believes that innovation is critical to its survival, and to reinforce that objective, it made innovation a KPI representing 20% of every team and individual’s performance review. In this episode of Leaders at All Levels, hosts Katherine W. Isaacs and Michele Zanini speak with Bidyut Dumra, group head of innovation and future of […]

• 1.AI and the Collapse of the www -- by Alex Chan

  This paper studies market design for generative AI intermediation. AI answer systems can improve user experience while diverting visits that finance publisher content and generate source-level quality signals. I show that an AI platform that underinternalizes future content reproduction retains too little referral traffic and can make costly open-web information subcritical, even with truthful content, accurate answers, and rational users. The mechanism can be self-reinforcing: less source-level measurement weakens conventional search, inducing further AI reliance. Sustainable repair requires replacing displaced revenue and deleted measurement through visitor-replacement royalties, audited provenance, human-information audits, and keystone-topic compensation.

• 2.A Practitioner's Guide to Using Large Language Models and Generative AI in Economic History -- by Andreas Ferrara

  Large language models (LLMs) are lowering the entry barriers to working with exciting data sources that used to require strong data science skills, such as handwritten ledgers, text, images, or sound recordings. This guide provides an introduction for researchers who are new to LLMs. It sets out a step-by-step workflow for turning a research idea into working code and data, and describes the four main ways of interacting with an LLM: the chat window, editor-integrated assistants, agentic coding tools, and the API. It then works through the decisions a practitioner meets in sequence, beginning with whether an LLM is the right tool and whether the data are allowed to be sent to one, then how to select models, write prompts, manage context limits, and control costs, and finally how to validate, reproduce, document, and correct LLM-generated ...

• 3.Compensation vs. Reinforcement: Experimental Identification of Parental Aversion to Inequality in Offspring -- by Felipe Barrera-Osorio, Leonardo Bonilla-Mejía, Matias Busso, Sebastian Galiani, Hyunjae Kang, Juan S. Muñoz-Morales, Juan Pantano

  Parents may invest differently across children by compensating the disadvantaged child or by reinforcing the child with higher expected returns. We study this question using a conditional cash transfer experiment that uniquely randomized transfers at the student level, generating exogenous variation in transfer exposure across siblings within the same household. The transfers increased short-run attendance among treated students but generated negative spillovers on untreated siblings: untreated siblings of treated students were 3.7 percentage points less likely to graduate from college, a decline of about 30 percent relative to the control mean. We interpret these effects using a dynamic model of household schooling decisions that identifies parental aversion to inequality in children’s educational outcomes. The estimated model implies li...

• 4.Prices Versus Quantities Revisited: What Do Policymakers Need to Know to Set Pigouvian Taxes and Subsidies? -- by Denise Dipasquale, Edward L. Glaeser, Adam M. Guren, Paul S. Willen

  What information do policymakers need to design Pigouvian taxes or subsidies? Standard logic suggests that it is sufficient to know the size of the externality and unnecessary to know about quantities. Yet this logic is incorrect if interventions have fixed costs, taxes create deadweight losses, or there are distributional concerns. We present a model in which these considerations can make it more valuable for policymakers to learn about equilibrium quantities. We apply the model to congestion pricing, which has high fixed costs, and to a proposed housing subsidy in Boston that features deadweight losses and distributional concerns.

• 5.Labor Market Responses to Tariffs: Frictions, Dynamics, and Policy Responses -- by Rafael Dix-Carneiro, Brian K. Kovak

  This article introduces the evidence and associated modeling frameworks contemporary economists use to understand the effects of trade and trade policy on labor markets, with a particular emphasis on labor-market frictions and adjustment dynamics. The effects of trade shocks differ across industries, regions, and occupations, implying the presence of important adjustment frictions in labor markets, and these effects evolve slowly over time, implying the need for dynamic frameworks rationalizing slow transitions. After reviewing the key insights from this literature, we discuss policies aimed at mitigating costs to workers and ensuring that the gains from trade are shared more equitably.

• 1.How Resilient Were Emerging Market Economies Through the 2022‑23 U.S. Monetary Tightening Cycle?

  The cross-border spillover effects of shifts in U.S. monetary policy have long been a focus of academics and policymakers alike. A common finding in the literature is that changes in the stance of U.S. monetary policy have sizable effects on economic activity and financial markets in emerging market economies (EMEs). In this post, we analyze one specific aspect of these spillovers: how EMEs fared through the U.S. monetary policy tightening cycle of 2022-23 relative to the predictions of a model, which was calibrated to capture empirically relevant features of these economies based on historical data. We find that more vulnerable EMEs fared better in both financial market and growth outcomes than would be expected from our model, while the relatively less vulnerable fared a bit better than the model predictions for financial outcomes but s...

• 2.The Post‑COVID Decline in the Labor Share

  The labor share of income in the U.S. is currently at its lowest-ever level in the post-war period. The labor share measures the fraction of economic output paid to workers as wages and salaries. As such, it is a useful benchmark for wage growth: when the labor share falls, it means that productivity, prices, or both are growing faster than wages. After much-studied drops in the 2000s, the labor share fell sharply again after the COVID pandemic. In this post, we compare the dynamics of the labor share post-COVID to earlier periods to understand whether the recent decline represents the continuation of a trend or a new and distinct phenomenon. We find that both the cyclicality of the labor share and the contribution of reallocation to the labor share post-COVID are similar to earlier periods.

• 3.Synthetic Stablecoins and Financial Stability

  On October 10, 2025, the announcement of a potential additional 100 percent tariff on Chinese goods drove risk-off moves across equities, Treasuries, credit spreads, and digital assets. Digital asset prices fell sharply, trading volumes surged, and liquidity vanished from key exchanges. In this post, we show how the price shock in digital assets was transmitted and amplified through a class of instruments called synthetic stablecoins—crypto assets whose structural design turned an external shock into a self-reinforcing deleveraging spiral within the crypto ecosystem.

• 4.The New York Fed DSGE Model Forecast—June 2026

  This post presents an update of the economic forecasts generated by the Federal Reserve Bank of New York’s dynamic stochastic general equilibrium (DSGE) model. We describe very briefly our forecast and its change since March 2026. To summarize, inflation forecasts are higher in 2026 than predicted in March. Projections for the short-run real natural rate of interest (r*) increased slightly relative to March.

• 5.The Unintended Effects of Interest Rate Caps: Credit Reallocation to Safer Borrowers

  Several states have recently capped consumer loan rates with the stated purpose of protecting borrowers. In a recent Staff Report, we study how these interventions have played out in three states. In our first post about that study, we showed that rate caps lead riskier borrowers to face rationing in the credit market. One question that naturally arises is what lenders do with the credit they used to provide to high-risk borrowers before the caps were imposed. Lenders that lend exclusively to high-risk borrowers (at rates above the cap) may decide to stop lending to high-risk borrowers in that state. Others, however, may ...

• 1.The Right Way to Tackle Developing Countries’ Cancer Crisis

  Some health advocates believe that weakening intellectual-property protections and capping prices for the most advanced cancer treatments would expand access in developing countries. But this well-intentioned argument overlooks an inescapable reality: without strong incentives for innovation, such treatments would not exist.

• 2.Enforce AI Guardrails Before It’s Too Late

  The Trump administration’s decision to impose export controls on Anthropic’s latest models is emblematic of the incoherent and shifting response from policymakers worldwide to the rapidly advancing AI sector. To ensure that these tools do not court disaster, governments must establish meaningful guardrails.

• 3.Is a European Single Market for Energy Such a Good Idea?

  Although textbook microeconomics shows why less energy-market fragmentation would be better for Europe as a whole, that does not mean it would be a good deal for countries producing low-carbon electricity at minimal marginal cost. In fact, if more integration creates free riders, the political backlash could be profound.

• 4.China’s Failed Rebalancing

  When China’s leaders first acknowledged the need to rebalance the economy nearly two decades ago, it seemed like a matter of when, not if. But with the household consumption share of Chinese GDP remaining stubbornly low, officials’ promises to boost domestic demand have lost all credibility.

• 5.Crowning a Healthy Populism

  Although recent political mobilizations in Hungary, Albania, and other parts of Central and Eastern Europe could be described as populist and nationalist, they have little in common with the aggressive ethnonationalism that MAGA professes. Could they be a model for a new form of politics?

• 1.China Mobile outlines ‘mobile intelligence’ vision at MWC Shanghai

  China Mobile has introduced AI into applications spanning sports, culture, and finance, while also exploring AI mobile phones, AI glasses, and humanoid robots In sum – what to know: Mobile intelligence – China Mobile said it plans to extend from…

• 2.BT and Verizon merge international arms in $4bn global enterprise JV

  BT and Verizon are combining international enterprise units into a 50/50 JV serving 3,000 customers across 180 countries, streamlining global connectivity services while letting both firms sharpen their focus on domestic markets and AI network demand. In sum – what…

• 3.Ericsson makes AI agents a first-class part of its OSS/BSS stack

  Ericsson’s architectural blueprint adds an agentic service experience layer In sum – what we know: Ericsson has put agentic AI at the center of its operations and business support stack, formalizing a new OSS/BSS architectural blueprint where AI agents are…

• 4.Subsea resilience needs to move beyond cable count – here’s why (Reader Forum)

  Subsea network resilience must be measured by corridor-level risk rather than cable count alone, argues Steve Roberts of EXA. He warns that shared dependencies, geopolitical instability, and repair constraints can undermine perceived route diversity. Route diversity has long been the…

• 5.Globe says AI execution, not adoption, is now the industry’s biggest challenge

  Speaking during a keynote session at MWC Shanghai 2026, Globe president and chief executive officer Carl Cruz said AI is fundamentally changing how companies compete In sum – what to know: AI execution focus – Globe chief executive officer Carl…

• 1.Source Error

  Check Feed

• 1.Toward AI-Native 6G Air Interface: A 3GPP Perspective on Protocol Framework

  Artificial intelligence (AI) is expected to play an important role in the sixth-generation (6G) air interface design, but making the air interface truly AI-native requires more than applying learning algorithms to individual radio functions. The deeper challenge is architectural: once AI influences how the user equipment and network interpret, predict, and adapt radio behavior, the air interface must provide common protocol semantics for coordinating such intelligence across vendors and deployments. This article presents a 3rd generation partnership project (3GPP) oriented perspective on the protocol framework for AI-native 6G air interface. We argue that standardization should preserve implementation freedom by avoiding prescription of model architectures, training methods, or model weights. Instead, 6G should define the protocol framewo...

• 2.Sequential and Generative Models for Vehicular Distributed MIMO Channel Prediction

  Vehicular communication is a key 6G use case requiring reliable and high-capacity connectivity under fast mobility and highly time-varying propagation conditions. However, large-scale vehicular channel estimation is costly and limited, impacting system-level performance of vehicular communications, and realistic channel prediction models are needed. This paper proposes a vehicular channel prediction framework based on real measured urban channels collected through a dedicated measurement campaign using the MaMIMOSA channel sounder. The framework enables the training and systematic benchmarking of sequential and generative models for both single-step and multi-horizon vehicular channel state information (CSI) prediction to assess prediction robustness across different forecasting horizons, including LSTM, TCN, a CNN-enhanced Transformer, a...

• 3.Lightweight PCGAE-Net: Parallel CrossGate Attention and Bottleneck AutoEncoder for Efficient 5G Channel Prediction

  Accurate channel state information (CSI) prediction is essential for proactive beamforming and resource management in 5G massive MIMO systems, yet the deployment of high-accuracy transformer-based predictors on base-station hardware remains challenging because the most capable models carry upwards of 30\,M parameters. This paper introduces Lightweight PCGAE-Net, which addresses the efficiency problem not by post-hoc compression but by correcting two architectural flaws in the current state of the art. The first is a sequential attention ordering bias: in CS3T-UNet, group-wise temporal attention (GTA) always operates on features that have already been transformed by cross-shaped spatial attention (CSA), distorting what temporal information GTA can capture. We remove this dependency by routing both attention modules to the same layer-normal...

• 4.Adaptive Machine Learning Framework for UAV Trajectory Optimization in O-RAN

  The deployment of unmanned aerial vehicles (UAV) as open radio units (O-RUs) in 6G cellular systems presents a promising opportunity to achieve scalable and adaptive network coverage. However, optimizing UAV trajectories in dynamic and unfamiliar environments remains a critical challenge, particularly due to the need for extensive retraining in each new scenario. In this paper, we introduce a novel UAV trajectory optimization framework that integrates enhanced continual transfer learning within the O-RAN architecture. The proposed system maintains a library of pre-trained models and employs a model selection mechanism to identify and transfer knowledge from the most relevant environments, minimizing adaptation time and improving efficiency. When no sufficiently similar model is available, a fallback model empowered by continuous refinemen...

• 5.WiWorld-RealData: A Real-World Multi-Modal Dataset for 6G Wireless World Models

  Wireless world models aim to represent, predict, and reason about wireless propagation by jointly understanding physical environments and channel responses. Realizing such models in sixth-generation (6G) digital twin channels requires datasets that capture measured wireless responses and environment states under real-world propagation conditions. This paper presents WiWorld-RealData, a real-world outdoor multi-band channel and multi-modal sensing dataset collected along campus mobile routes. WiWorld-RealData provides measured channel impulse responses (CIRs) at 3.7 GHz and 6.775 GHz, together with multi-view images, panoramic images, light detection and ranging (LiDAR) point clouds, millimeter-wave (mmWave) radar records, and global navigation satellite system (GNSS) trajectories. Through unified file organization and metadata manifests, ...

• 1.(In)Efficient Market States and Rough Volatility Detected via Grunwald-Letnikov Fractional Derivative

  Testing self-similarity in fractional processes from a single observed trajectory is difficult under long-range dependence, because the associated Kolmogorov--Smirnov (KS) statistic undergoes a phase transition when $H>1/2$. In this regime, the classical limit collapses to a non-functional absolute Gaussian law and finite-sample convergence becomes severely distorted. This paper introduces a regime-adaptive KS/GL--KS framework based on the discrete Grünwald--Letnikov (GL) fractional derivative. The GL filter removes the low-frequency long-memory singularity while preserving the finite-dimensional $H$-self-similarity needed for distributional identification. We derive the filtered empirical-process limit, prove consistency and local asymptotic behavior of the resulting Hurst estimator, and validate the method through Monte Carlo simulat...

• 2.CryptoGAT: Are Time Series Models Effective for Cryptocurrency Forecasting?

  Summary available at source link.

• 3.The Inference-Compute Frontier and a Latency-Efficient Architecture for Limit Order Book Prediction

  We study whether a scaling-law-style inference-compute frontier appears in limit order book prediction. Using FI-2010 and a suite of models ranging from small decision trees to neural LOB architectures, we find that the realized empirical frontier of predictive loss versus structural forward work is well summarized by a power law. In particular, with MLPLOB held out as an architecture family, a power-law fit to the low- and mid-compute non-MLPLOB frontier extrapolates across multiple orders of magnitude and attains $R^2=0.941$ on the excluded high-compute MLPLOB target frontier. A similar exercise in latency space gives substantially weaker results, showing that latency is not merely noisy compute. We use this gap to motivate FastBiNLOB, a dense axis-separable LOB mixer built from hardware-friendly temporal and feature mixing operations...

• 4.Time-dependent weighted directed networks of cryptocurrency interaction from high-frequency returns

  We investigate the evolving structure of interactions in cryptocurrency markets using a network-based framework constructed from high-frequency price data spanning 2020-2025. Directed and weighted networks are constructed from statistically significant Granger causal relationships between cryptocurrency log-returns, enabling us to quantify the flow of influence across assets. We find that normalized returns exhibit heavy-tailed distributions, consistent with the presence of large intermittent fluctuations and in line with stylized facts of financial markets. The resulting networks display pronounced heterogeneity in link weights and nodal strengths, indicating that a small subset of cryptocurrencies contributes disproportionately to market dynamics. By ranking cryptocurrencies based on their nodal out-strength, we uncover a dynamically ev...

• 5.Multi-Stream Temporal Fusion for Financial Fraud Detection

  Financial fraud detection in digital banking requires reasoning over multiple heterogeneous event streams -- transactions, login sessions, risk signals -- that individually appear benign but collectively reveal fraudulent patterns. We propose the Multi-Stream Fraud Transformer (MSFT), a unified architecture that encodes each event stream with independent Transformer encoders and fuses their representations through configurable mechanisms. We conduct a systematic ablation study comparing five fusion strategies: concatenation, gated fusion, time-aware positional encoding, cross-stream attention, and a full combination. On a large-scale dataset (10M users, 1.5% fraud rate) with 85M parameter models, we demonstrate that (1) sequence models significantly outperform gradient-boosted trees operating on aggregated features (0.74 vs. 0.99 AUROC), ...

• 1.Toward AI-Native 6G Air Interface: A 3GPP Perspective on Protocol Framework

  Artificial intelligence (AI) is expected to play an important role in the sixth-generation (6G) air interface design, but making the air interface truly AI-native requires more than applying learning algorithms to individual radio functions. The deeper challenge is architectural: once AI influences how the user equipment and network interpret, predict, and adapt radio behavior, the air interface must provide common protocol semantics for coordinating such intelligence across vendors and deployments. This article presents a 3rd generation partnership project (3GPP) oriented perspective on the protocol framework for AI-native 6G air interface. We argue that standardization should preserve implementation freedom by avoiding prescription of model architectures, training methods, or model weights. Instead, 6G should define the protocol framewo...

• 2.Sequential and Generative Models for Vehicular Distributed MIMO Channel Prediction

  Vehicular communication is a key 6G use case requiring reliable and high-capacity connectivity under fast mobility and highly time-varying propagation conditions. However, large-scale vehicular channel estimation is costly and limited, impacting system-level performance of vehicular communications, and realistic channel prediction models are needed. This paper proposes a vehicular channel prediction framework based on real measured urban channels collected through a dedicated measurement campaign using the MaMIMOSA channel sounder. The framework enables the training and systematic benchmarking of sequential and generative models for both single-step and multi-horizon vehicular channel state information (CSI) prediction to assess prediction robustness across different forecasting horizons, including LSTM, TCN, a CNN-enhanced Transformer, a...

• 3.Adaptive Machine Learning Framework for UAV Trajectory Optimization in O-RAN

  The deployment of unmanned aerial vehicles (UAV) as open radio units (O-RUs) in 6G cellular systems presents a promising opportunity to achieve scalable and adaptive network coverage. However, optimizing UAV trajectories in dynamic and unfamiliar environments remains a critical challenge, particularly due to the need for extensive retraining in each new scenario. In this paper, we introduce a novel UAV trajectory optimization framework that integrates enhanced continual transfer learning within the O-RAN architecture. The proposed system maintains a library of pre-trained models and employs a model selection mechanism to identify and transfer knowledge from the most relevant environments, minimizing adaptation time and improving efficiency. When no sufficiently similar model is available, a fallback model empowered by continuous refinemen...

• 4.WiWorld-RealData: A Real-World Multi-Modal Dataset for 6G Wireless World Models

  Summary available at source link.

• 5.Integrated Sensing and Communications for Real-time Avatar Control in XR over 5G

  Extended Reality (XR) presents a challenging use case for 5G and 6G networks, requiring high data-rates and lowlatency communication to deliver a truly immersive experience. Moreover, in order to seamlessly translate physical actions to the virtual world, accurate gesture recognition and pose estimation are required. Current XR interaction solutions based on handheld controllers and cameras cannot easily capture full-body poses, inhibit the free use of hands, and require good visibility and a clear line of sight. In this work, we propose a multimodal sensing architecture for XR that combines 5G MillimeterWave (mmWave) Integrated sensing and communication (ISAC) and surface electromyography (sEMG) signals. 5G mmWave ISAC cannot only be used to deliver content wirelessly to the Head-mounted display (HMD), but also the same communication sig...

• 1.V-TSN: A Software-Defined TSN Overlay for General-Purpose Networks

  Time-Sensitive Networking (TSN) extends Ethernet with deterministic communication for time-critical applications such as industrial automation, in-vehicle networks, and cyber-physical systems. However, realizing TSN behavior without dedicated hardware is difficult. During design and validation, offline simulation cannot run application software at real-time speed when costly specialized TSN hardware is not (yet) available. At deployment time, many systems run on general-purpose and cloud networks with no native TSN support, where provisioning full TSN hardware is unnecessary or impractical for applications that tolerate relaxed timing. In this paper, we introduce Virtual Time-Sensitive Networking (V-TSN), a software-defined overlay that realizes gPTP-based synchronization and TSN traffic shaping over general-purpose, non-deterministic net...

• 2.Toward AI-Native 6G Air Interface: A 3GPP Perspective on Protocol Framework

  Artificial intelligence (AI) is expected to play an important role in the sixth-generation (6G) air interface design, but making the air interface truly AI-native requires more than applying learning algorithms to individual radio functions. The deeper challenge is architectural: once AI influences how the user equipment and network interpret, predict, and adapt radio behavior, the air interface must provide common protocol semantics for coordinating such intelligence across vendors and deployments. This article presents a 3rd generation partnership project (3GPP) oriented perspective on the protocol framework for AI-native 6G air interface. We argue that standardization should preserve implementation freedom by avoiding prescription of model architectures, training methods, or model weights. Instead, 6G should define the protocol framewo...

• 3.Dependency-Aware Dominant Resource Fairness for Multi-Tenant Multi-Resource Systems

  Multi-resource allocation in network-congested, multi-tenant systems in which demand exceeds available capacity is challenging, as there is no straightforward way to determine how much of each resource to assign, especially when resources are interdependent. Classical approaches such as Dominant Resource Fairness (DRF), which generalizes Max-Min Fairness (MMF) to multiple resources, assume linear proportional dependencies across resources, requiring allocations to follow fixed proportions implied by tenants demands. However, this assumption may lead to inefficient allocations and resource waste, with allocated resources that go unused in practice. In this paper, we consider a multi-resource orchestrator and propose the Dependency-aware Dominant Resource Fairness (DDRF) policy, a centralized generalization of DRF that considers inter-resou...

• 4.RQ-SAFE: Coupled Request-Resource Scheduling for Online Edge SFC-DAGs

  Intent-driven edge services allow multiple virtual network function (VNF) segments in a service function chain directed acyclic graph (SFC-DAG) to be locally reordered without changing service semantics, creating richer request-side orchestration freedom. Existing orchestration methods mainly optimize VNF placement, routing, or queue-aware scheduling for a predetermined service order; they do not fully exploit this freedom or couple it with runtime resource scheduling. This paper presents RQ-SAFE, a request-resource coupled scheduling framework for online edge SFC-DAG orchestration with checked commitment. RQ-SAFE evaluates each feasible local order by previewing its resource-side consequences on the current edge infrastructure, and uses the retained order to guide VNF instance selection and path construction. Queue state is used througho...

• 5.BRAVR: An AP-Assisted Online DRL Mechanism for Interactive VR Bitrate Adaptation over Wi-Fi

  Interactive virtual reality (VR) streaming over Wi-Fi requires stringent latency and reliability guarantees, which become increasingly difficult to achieve under dynamic channel conditions and shared medium contention. These challenges make real-time bitrate adaptation a critical yet fundamentally difficult control problem, particularly under limited visibility of the underlying network conditions. This paper formulates VR bitrate adaptation as a network-aware, online decision-making problem and proposes BRAVR, a decentralized deep reinforcement learning (DRL) mechanism designed to optimize visual quality while maintaining streaming performance and promoting airtime fairness in multi-user scenarios. BRAVR integrates application-layer observations with lightweight wireless network statistics collected at the Wi-Fi access point (AP) serving...