AI radar · Pulse

Radar

The AI stack

Hugging Face models, datasets, and spaces ranked by a likes + downloads heat score. arXiv cs.AI is the research tape. Counts come from public APIs — if a field is missing, it is omitted.

As of Oct 21, 5:25 AM EDT · revalidates daily

Sources live · hf-models 24 · hf-datasets 24 · hf-spaces 24 · arxiv 8

Models

Hottest weights

Data

Datasets

Research

arXiv cs.AI

2026-10-20 · Noam Rotstein, Yoad Tewel, Chen Tessler
Joints, Camera, Action! Co-Generating 3D Motion and Video for Arbitrary Rigs
Video generation models can now depict remarkably realistic motion. Yet it is unclear how to use this motion beyond pixels: to animate a specific 3D asset and replay the animation in its 3D scene. For instance, a video model can show a monkey jumping over a crossbar, but it canno
2026-10-20 · Yazhuo Liu, Kunqing Ding, Xiaoguo Gong
GBClaw: An agentic AI framework for atomistic grain boundary research and mechanistic discovery
Agentic artificial intelligence (AI) systems hold immense potential for automating complex workflows in materials discovery and design, but extending them to domain-intensive mechanistic investigations remains challenging. Here, we present GBClaw, an agentic AI framework for inve
2026-10-20 · Rahul Sajnani, Diego Romeres, Mingyuan Zhang
UniWALM: Unified World Action Language Model for Embodied Robot Control
Motivated by the human ability to integrate experience across language, perception, and action, we introduce UniWALM, a unified world-action-language model that learns video, action, and language in a single shared representation, where each modality is both a conditioning signal
2026-10-20 · Jad Yammine, Malek Wahidi, Abrar Anwar
Industrial ReMEmbR: Scaling Spatiotemporal Memory to Robot Fleets

Repos

Language mix on the GitHub tape

06

nvidia/PhysicalAI-Autonomous-Vehicles

dataset · 1.2k likes · 262k dl

82

07

Anthropic/hh-rlhf

dataset · 2.4k likes · 35k dl

80

08

roneneldan/TinyStories

dataset · 1.2k likes · 144k dl

80

09

XiaomiMiMo/MiMo-V2.6-RL-oss

dataset · 989 likes · 195k dl

80

10

tatsu-lab/alpaca

dataset · 1.4k likes · 83k dl

79

11

HuggingFaceH4/ultrachat_200k

dataset · 1.2k likes · 91k dl

79

12

OpenAssistant/oasst1

dataset · 1.6k likes · 42k dl

78

13

ILSVRC/imagenet-1k

dataset · 1.0k likes · 92k dl

78

14

databricks/databricks-dolly-15k

dataset · 1.4k likes · 40k dl

77

15

gsdf/EasyNegative

dataset · 1.2k likes · 37k dl

76

16

Open-Orca/OpenOrca

dataset · 1.6k likes · 13k dl

75

17

tiiuae/falcon-refinedweb

dataset · 975 likes · 36k dl

75

18

FreedomIntelligence/medical-o1-reasoning-SFT

dataset · 1.2k likes · 22k dl

74

19

cais/hle

dataset · 1.0k likes · 29k dl

74

20

bigcode/the-stack

dataset · 1.1k likes · 7.4k dl

70

21

lmsys/lmsys-chat-1m

dataset · 1.1k likes · 4.7k dl

69

22

togethercomputer/RedPajama-Data-1T

dataset · 1.2k likes · 3.1k dl

68

23

allenai/dolma

dataset · 1.1k likes · 2.9k dl

68

24

Nerfgun3/bad_prompt

dataset · 987 likes · 3.7k dl

68

Apps

Spaces

Warehouse robot fleets continuously observe operational conditions, but preserving these observations as reusable semantic memory incurs substantial computation, communication, and storage costs. For example, explaining recurring aisle blockages can require reviewing video from m

2026-10-20 · Yanbo Wang, Siqi Luo, Siyu Zhang

MindView: Benchmarking Generative Multi-View Spatial Intelligence

Spatial intelligence has recently been assessed through generative tasks such as spatially grounded image editing, yet language-guided generation of spatially consistent views from partial multi-view observations remains underexplored. We introduce MindView, a benchmark for gener

2026-10-20 · Emile Anand, Mohammed Rehan Parwani, Xinyuan Cao

Adaptive Multi-Token Prediction: How far should we go?

Multi-token prediction (MTP) can accelerate language-model decoding by drafting several tokens and verifying them in a single target-model pass. However, the best speculation horizon changes with context: longer drafts offer greater amortization when reliable, but can waste compu

2026-10-20 · Jitendra Solanki, Anand Todkar, Ziran Min

Semantic Observability for Autonomous Systems Diagnostics

Modern distributed microservice deployments generate massive volumes of heterogeneous, multi-format observability logs across physically and logically partitioned infrastructure. Isolating root causes during cascading failures remains an arduous task for Site Reliability Engineer

2026-10-20 · Noor Abo Mokh, Teresa Lynn, Nizar Habash

Benchmarking Arabic Dialects for Math Problems

Mathematical reasoning remains underexplored for low-resource languages, including Dialectal Arabic. To bridge this gap, we introduce the DAR-MGSM (Dialectal Arabic Multilingual Grade School Math) Benchmark, the first manually curated benchmark for evaluating mathematical reasoni

Joints, Camera, Action! Co-Generating 3D Motion and Video for Arbitrary Rigs

GBClaw: An agentic AI framework for atomistic grain boundary research and mechanistic discovery

UniWALM: Unified World Action Language Model for Embodied Robot Control

Industrial ReMEmbR: Scaling Spatiotemporal Memory to Robot Fleets

MindView: Benchmarking Generative Multi-View Spatial Intelligence

Adaptive Multi-Token Prediction: How far should we go?

Semantic Observability for Autonomous Systems Diagnostics

Benchmarking Arabic Dialects for Math Problems