WAM‑Sim0.1 is a driving world model that uses a persistent hierarchical memory to anchor context across pose, scene, and instance tokens, addressing structural, identity, and quality drift in autoregressive simulations. It incorporates online flow‑matching reinforcement learning to refine self‑generated rollouts, guided by visual rewards and a vision‑language physical judge, thereby improving visual sharpness and physical plausibility. On the nuScenes benchmark, WAM‑Sim achieves high generation fidelity (FID = 3.47, FVD = 25.72) and strong downstream perception metrics (24.16 mAP, 71.23 road mIoU), while maintaining scene similarity and identity coherence over minute‑scale rollouts.
By Jiahao Cui, Jiaye Li, Wenhao Zhang, Yaoyi Li, Shan Luan, Quanhui Tang, Jianhua Han, Lin Cheng, Hang Xu, Siyu Zhu
MITO is the first millimeter‑wave dataset featuring diverse everyday objects, collected with a UR5 robotic arm equipped with two mmWave radars and an RGB‑D camera. It includes over 24 million mmWave frames, 550 high‑resolution mmWave images in both line‑of‑sight and non‑line‑of‑sight conditions, along with RGB‑D images, segmentation masks, and raw signals for 76 objects. An open‑source simulator is provided to generate synthetic mmWave images from any 3D triangle mesh, and benchmarks for NLOS segmentation and classification are established to showcase the dataset’s utility.
By Tara Boroushaki, Laura Dodds, Cusuh Ham, Fadel Adib
OpenAI introduces BioBasicsBench, a 100-task benchmark measuring how well AI agents interpret raw experimental data and handle everyday wet-lab science.
ML code should run where researchers choose, not where it was first written. Today, machine learning code mainly lives in two ecosystems, PyTorch on NVIDIA GPUs and JAX on Google TPUs. Migrating repos...
Coding agents are among the most economically consequential deployments of LLMs to date, making reliable evaluation increasingly important. Yet even benchmarks that are carefully validated at release...
The growing use of AI in mathematics raises a question: can we detect AI involvement in proofs, separately from AI assistance with writing? Historical arXiv papers, recent papers disclosing substantiv...
Self-consistency (SC) is a widely used decoding strategy for improving chain-of-thought reasoning, but it requires sampling many complete reasoning trajectories with a preset budget. Its efficient var...
Online 3D bin packing with a processing buffer requires choosing which box to pack, its orientation, and its placement while keeping the stack stable under unknown future arrivals. A collapse ends the...
Personal AI agents face social engineering attacks supported by forged multimodal content, but existing evaluations mainly focus on text-based dialogue, leaving their intrinsic security under such att...
High-resolution tropical cyclone surface wind speeds derived from synthetic aperture radar (SAR) observations can support storm structure and intensity modelling, but remain spatially and temporally s...
Multi-modal large language models (MLLMs) with impressive perception and understanding capabilities will enable a range of downstream applications. However, their substantial computational and memory...
Keyword retrieval returns expressions of personal anxiety together with many other uses of anxiety-related language. We present a benchmark of 1,000 Arabic Twitter/X posts retrieved with the keyword q...
Large-scale identity-labeled face datasets are essential for training modern face recognition models, yet collecting and annotating such data is costly and raises privacy and ethical concerns. Existin...
The growing adoption of digital pathology has rapidly expanded collections of hematoxylin-and-eosin whole-slide images, yet efficient navigation through these archives remains challenging, as metadata...
Accurate 4D reconstruction remains challenging: jointly estimating camera motion, scene geometry, and non-rigid deformation entails highly nonconvex optimization, making global optimality difficult to...
Large language models use auxiliary conditioning to incorporate additional information into time-series forecasts. However, auxiliary conditioning does not consistently improve accuracy and can increa...
LLM-based recommender systems increasingly generate chain-of-thought (CoT) text before each item prediction to activate the LLM's reasoning capacity, costing tens to hundreds of extra tokens per recom...
Conditional generative surrogates in high-energy physics, such as models that predict reconstructed jets from partons, are often evaluated on $1\text{D}$ marginal divergences among other metrics. Take...
Volatility dynamics can carry information about temporal relationships that is not apparent in the conditional mean. Extracting this information is difficult because the persistence of each variable c...