Provides a dual-path approach for spatial vision-language models: a Language-Only Reasoning (LOR) path for stepwise linguistic deduction and a Detect-Then-Reason (DTR) path that detects 3D cues via region tokens before numerical inference. Trains with chain-of-thought cold-start supervision and reinforcement learning to improve 3D grounding and multi-step spatial reasoning.
Evaluates multimodal LLMs' ability to reconstruct past observations and act in controllable non-Markov games. Introduces RNG-Bench with two games (Matching Pairs, 3D Maze), three controllable difficulty axes, a head-to-head duel protocol, and a Memory Gap metric to separate forgetting from action errors.
AI Image2026
Generates images from natural-language prompts as an 8-step distilled checkpoint of Krea 2, optimized for fast iterative text-to-image workflows with style references and 1K–2K resolution outputs.
AI Model2026
Performs one-shot, long-horizon OCR and document parsing by using Reference Sliding Window Attention (R-SWA) to keep the decoder KV cache constant, enabling single-pass multi-page transcription; code, model weights and an accompanying arXiv report are provided.
AI Model2026
Provides GGUF-quantized weights and runtime assets for running the Qwythos-9B reasoning LLM locally via llama.cpp and compatible runtimes. Key features include 1,048,576-token YaRN long-context, native function-calling, multimodal image input (requires mmproj), and multiple quantization/MTP variants tuned for different size/quality tradeoffs.
AI Dataset2026
Zeren Jiang, Yushi Lan +9Visual Geometry Group, University of Oxford, Nanyang Technological University +1
Provides multiview synthetic RGB video clips with per-frame depth, instance masks, dense long-range 3D point tracks, camera poses, and SMPL‑X human pose/shape labels for 4D reconstruction, tracking, and geometry-aware novel-view synthesis. Includes ~4.7K clips (1.4M frames) and is licensed for AI training.
AI Dataset2026
Provides 600,000+ first-person player-round videos (10,000+ hours) with per-frame keyboard, mouse-delta, and 3D trajectory annotations in WebDataset shards—built for training world models, action-conditioned video, and imitation-learning workflows (non-commercial license).
Provides a rubric-based benchmark that converts dense image captions into instance-specific atomic checks (Must-Right and Easy-Wrong) and a gated scoring rule, aiming to expose perceptual brittleness and better align multimodal model evaluation with human judgment.
AI Dataset2026
Provides synchronized four-perspective Rocket League match recordings with per-frame H.264 video, player action streams, event logs, and privileged physics state — released as WebDataset shards in a ~4,000-hour slice (1,000 match-hours × 4 perspectives). Includes 720p@20fps video, multi-hot keyboard actions, and CC BY-NC-SA-4.0 license.
AI Dataset2026
Provides ~494.7 hours of trimmed native PC/console gameplay screen recordings organized by game, with per-session clips plus input and per-frame event annotations. Each workflow includes clip.mp4, events.json, frame_events.json, and metadata — suitable for training vision-action, behavior-cloning, and gameplay understanding models.
AI Dataset2026
Provides 2,165 historical natural-history page scans paired with ~99.95% expert transcriptions and pixel-aligned PAGE XML layout ground truth for OCR and layout evaluation; multilingual (EN/FR/DE/LA), CC-BY 3.0.
AI Dataset2026
Provides ~50M multimodal annotations organized for unified training across structured visual understanding, segmentation, dense geometric prediction, and multi-view reconstruction — released as task-specific JSONL files that reference original image assets rather than redistributing raw images.