Embodied AI2026
Generates robot manipulation actions from visual observations and text instructions using a 1.5B vision-language-action model. Uses streaming context and visual-token compression to cut per-step compute, runs a unified policy across tasks, and is open-sourced on Hugging Face under Apache-2.0.
Embodied AI2026
Predicts eight future [x,y,yaw] waypoints for language-conditioned embodied person-following using fused DINOv3 and SigLIP visual features; trained with quality-driven, DAgger-style self-evolving data and optimized for on-device inference (~5+ FPS, ~180 ms).
AI Video Papers2026
Predicts variable-cardinality sets of evidence intervals in videos to temporally ground queries using multimodal large language models. Combines caption-derived multi-span supervision, a temporal Wasserstein matching-free reward, and temporal IoU, yielding strong mIoU gains across multiple benchmarks.
AI Model2026
GGUF-quantized releases (NEO IMATRIX + MTP) of a multi-stage fine-tuned, uncensored Qwen3.5-9B model with vision enabled and a native 256k context window—optimized for instruction following, reasoning and image-text-to-text workflows; released under Apache-2.0.
AI Dataset2026
Provides 2,000 hours of synchronized, high‑fidelity robot‑free bimanual manipulation demonstrations with multi‑view video, calibrated end‑effector trajectories, gripper states, and language annotations. Curated from a 20,000+ hour corpus; features 6 camera views, ~3 mm pose accuracy, <40 µs cross‑sensor sync, and LeRobot v3‑style Parquet+MP4 export under CC BY 4.0.
AI Video Papers2026
Personalizes subject-driven videos to preserve human identity and accurate human–object interactions by integrating multimodal references and MLLM-derived semantics. Introduces global multimodal guidance in self-attention and modality-reference embeddings to align MLLM features with VAE tokens, supporting both inter- and intra-subject inputs (e.g., OCR, multi-view).
AI Video Papers2026
Generates interactive long-horizon 24-fps video worlds (540p/720p) from text, image, or video inputs. Uses a 15B video diffusion transformer with a bounded visual context (sink frame, compressed temporal history, geometry-aligned spatial memory, recent-frame conditioning) and a discrete autoregressive distillation that cuts inference to ~4 sampling steps per chunk.
AI Model2026
Adds vision to GLM-5.2 by attaching a MoonViT encoder and a trained 49.5M-parameter PatchMerger projector to enable image→text multimodal reasoning; text and vision backbones are frozen, uses NVFP4 quantized weights and targets Blackwell B200 GPUs.
AI Model2026
Generates and reasons about multimodal physical-world content—text, images, video and action trajectories—conditioned on text, images, video and robot/vehicle action inputs. An edge-sized (4B) Mixture‑of‑Transformers omni-model optimized for single‑GPU inference and Physical AI tasks (image→video, action prediction, robot policy).
Synthesizes RGB frames from structured world states exported by physics engines; it reformulates a heavy generative renderer into a few-step autoregressive streaming model and uses lightweight distilled codecs to reach playable ~30 FPS while preserving G-buffer and prompt control.
Efficient 4B-scale image generation and editing model family that pairs a lightweight VAE tokenizer (Mage-VAE) with a native-resolution multimodal diffusion transformer, reducing tokenization cost by an order of magnitude and enabling few-step high-resolution generation and editing.
AI Model2026
Efficient 4B native-resolution diffusion foundation model for text-to-image generation and instruction-based image editing. Uses a lightweight Mage‑VAE tokenizer and a 4B NR‑MMDiT backbone to produce 512–2048 outputs with low memory and fast inference; ships in Base, RL-aligned and few-step Turbo variants.