11
Mini Amusement Parks (MAPs): A Testbed for Modelling Business Decisions
Researchers have developed a new simulation testbed designed to evaluate how AI models handle complex, long-horizon business decision-making.
11
EquiFusion: Kinematics-Agnostic Human Motion Prediction via Equivariant Latent Diffusion
EquiFusion is a new kinematics-agnostic latent diffusion model that enables human motion prediction without being restricted to specific skeleton structures.
11
The Nuts and Bolts of Natural Language to SQL Translation: A Systematic Analysis of Model Pipeline Optimisation Approaches and their Interactions
This study evaluates various pipeline optimizations for natural language to SQL translation to improve the efficiency and performance of lightweight models.
11
Distributed Agent System: Fault-Tolerant Collaboration Among Embodied Agents
The Distributed Agent System framework is introduced to enhance fault tolerance and collaboration among embodied AI agents across device, edge, and cloud environments.
11
Explainable Geospatial AI for Satellite Ground Station Siting Using LiDAR-Derived Terrain Intelligence
Researchers developed an explainable AI model using LiDAR data to improve the accuracy of satellite ground station siting by better accounting for local terrain obstructions.
11
Learning to Fine-tune Foundation Models under Resource Limitations
This research explores optimal strategies for continually fine-tuning foundation models on resource-constrained devices by balancing compute costs against model performance.
11
Transformer-Guided Swarm Intelligence for Frugal Neural Architecture Search
A new neural architecture search framework utilizes swarm intelligence and transformer controllers to enable efficient model design on consumer-grade hardware.
11
Concept-Guided Spatial Regularization for World Models in Atari Pong
Researchers evaluated the performance of five prominent world-model agents in Atari Pong to better understand their isolated capabilities.
11
World Models as Adversaries: Multi-Agent Self-Play Fine-Tuning for Robust Motion Planning
A new adversarial training method for autonomous vehicle motion planning uses multi-agent self-play to improve robustness in rare, safety-critical traffic scenarios.
11
Motif: Discovering and Automating Personal Web Workflows
Motif is a new system that passively observes user behavior to identify and automate repetitive web-based workflows.
11
Multi-Turn On-Policy Distillation with Prefix Replay
The proposed Replayed-Prefix On-Policy Distillation method aims to reduce the computational costs associated with training agentic LLMs by using off-environment data.
11
Pipette: An Embodied Simulation Platform, Benchmark, and Data-Efficient Augmentation Framework for Wet-Lab Robotics
Pipette is a new simulation platform and benchmark created to improve the training and data efficiency of robotics systems used in biomedical laboratory environments.
11
Structured Thoughts For Improved Reasoning And Context Pruning
A new framework called Structured Thoughts organizes large language model reasoning into distinct blocks to enhance efficiency and context management.
11
Multi-Scale ViT Inference with Habitat-Fit Priors and kNN Retrieval for Multi-Species Plant Identification
A multi-scale vision transformer approach is detailed for identifying multiple plant species in high-resolution photographs using limited training data.
11
ARMOR: Stabilizing On-Policy LLM RL with Off-Policy Anchor Samples
The ARMOR framework introduces off-policy anchor samples to stabilize reinforcement learning in large language models and prevent over-optimization.
11
Towards Objective Dysgraphia Detection: A Multi-Branch Deep Learning Approach for Online Handwriting Analysis
Researchers developed a deep learning-based approach to analyze online handwriting for more objective and efficient dysgraphia detection in children.
11
MG$^2$-RAG: Multi-Granularity Graph for Multimodal Retrieval-Augmented Generation
MG2-RAG is a proposed multimodal retrieval-augmented generation framework that uses multi-granularity graphs to improve reasoning and preserve fine-grained visual data.
11
Automated identification of Ichneumonoidea wasps via YOLO-based deep learning: Integrating HiresCam for Explainable AI
Researchers have developed a deep learning framework using YOLO and explainable AI techniques to automate the taxonomic identification of parasitoid wasps.
11
Beyond Bayesian Nash: Learning Minimax-Regret Equilibria for Adversarial Team Games under Asymmetric Information
This research introduces a method for learning minimax-regret equilibria in adversarial team games with asymmetric information to counter deceptive opponent strategies.
11
The Cross-Architecture Substrate: A Domain-Transcendent, Calibration-Surviving Geometric Invariant of Modern Vision Encoders
A study reveals that diverse modern vision encoders converge toward a shared sixteen-dimensional geometric structure, termed the cross-architecture substrate, regardless of their training objectives.
11
Physics-Informed Structure Anchoring With Capture-Aware Prototype Calibration for Cross-Environment RF Fingerprinting
This study proposes a physics-informed framework to improve the robustness of radio frequency fingerprinting models across changing physical environments.
11
Enabling Agents to Communicate Entirely in Latent Space
Researchers have introduced a communication framework that allows AI agents to exchange information directly through latent states rather than relying on natural language tokens.
11
Webyne plots 100MW footprint across India
Webyne is planning to expand its data center footprint in India to 100MW, significantly increasing its current 10MW capacity.
11
韩国综指跌幅扩大至5%
Major South Korean semiconductor manufacturers SK Hynix and Samsung Electronics experienced significant stock price declines during a broader market downturn.