Papers

TokenGate: Learnable Token Pruning for Fast, Memory-Efficient LLM Prefill

TokenGate learns a residual scorer at a middle layer, retains top-K document tokens plus request scaffolding, and uses token-gated attention to train through the discrete selection. It reports 1.4-11.3x faster TTFT and 1.5-3.4x lower peak prefill memory on Qwen3 and Qwen3.5 without accuracy loss; robotics performance remains untested.

Large Language ModelsToken Pruning

Liana Patel, Rishabh Ranjan, Ion Stoica · 306 views · Oct 19, 2026

Training Generalist Navigation Policies via Action Self-Distillation

Action Self-Distillation turns future trajectory feedback into an inferred language goal, then trains one open-source vision-language model to predict reasoning and ego-frame paths from onboard observations without human language or reasoning labels.

NavigationVLA

Bowen Feng, Keqin Wang, Yulong Yang · 302 views · Oct 13, 2026

CanonNav: Disentangling Navigation Behavior from Camera Geometry in Cross-Platform Visual Navigation

While visual navigation has advanced through imitation learning from cross-platform demonstrations, fully leveraging such data remains challenging. First, directly learning from image-trajectory pairs entangles navigation behavior with platform-dependent camera geometry. This hinders consistent learning by forcing the policy to implicitly infer camera geometry from visual observations, an inherently ill-posed problem. Second, imitation learning from demonstrated trajectories captures the expert's chosen motion but leaves the intermediate decisions underlying that motion implicit. To address these issues, we propose CanonNav, a visual navigation framework that disentangles navigation behavior from camera geometry and incorporates complementary planning supervision into learning from cross-platform demonstrations. CanonNav introduces camera geometry canonicalization, which transforms visual observations and trajectories into a camera-consistent representation space. Building on this representation, we derive safety and local-progress supervision using pseudo-labels from an offline traversability estimator. Safety supervision penalizes unsafe trajectories, while local-progress supervision guides where the robot should advance. Experiments across diverse camera configurations and environments show that, despite using only RGB at inference, CanonNav consistently outperforms RGB-based baselines and even surpasses RGB-D-based methods in challenging scenarios.

Visual navigationCross-Platform

Dong-Wook Kim, Ji-Hoon Hwang, E-In Son · 210 views · Aug 31, 2026

Draft: A Parametric Tool for Robot Design Exploration

Robot performance is often limited by the cost of iterating on morphology and control together, since every computer-aided design (CAD) change has to be carried into a simulation-ready model before control work begins. Co-design methods attempt to close this gap, but each uses a model generator written for a single platform or lack the use of real-world data to suggest that designs are plausible. We present Draft, a parametric generation tool whose generalized engine compiles any parametric tree of serial chains into a simulation-ready MJCF model, without CAD. It allows engineers to explore design tradeoffs through easily adjustable models and evaluate how changes influence controller performance. Draft grounds the free parameters of each design using trends fitted to a survey of $114$ actuators and $49$ published robot descriptions, so that a generated robot is anchored to real-world hardware. We validate those trends wholistically by building twins of four off-the-shelf robots, whose masses agree to 1.10× geometric mean fold error. Finally, we demonstrate how Draft exposes design tradeoffs by evaluating three quadrupeds through a two-stage reinforcement learning curriculum.

Robot DesignParametric Modeling

David Nguyen, Marcelo Coelho, Sangbae Kim · 187 views · Sep 29, 2026

HA-VLN 2.0: An Open Benchmark and Leaderboard for Human-Aware Navigation in Discrete and Continuous Environments with Dynamic Multi-Human Interactions

Vision-and-Language Navigation (VLN) has been studied mainly in either discrete or continuous spaces, with little attention to dynamic, crowded environments. We present HA-VLN 2.0, a unified benchmark introducing explicit social-awareness constraints. Our contributions are: (i) a standardized task and metrics capturing both goal accuracy and personal-space adherence; (ii) HAPS 2.0 dataset and simulators modeling multi-human interactions, outdoor contexts, and finer language-motion alignment; (iii) benchmarks on 16,844 socially grounded instructions, revealing sharp performance drops of leading agents under human dynamics and partial observability; and (iv) real-world robot experiments validating sim-to-real transfer, with an open leaderboard enabling transparent comparison. Results show that explicit social modeling improves navigation robustness and reduces collisions, underscoring necessity of human-centric approaches. By releasing datasets, simulators, baselines, and protocols, HA-VLN 2.0 provides a strong foundation for safe, human-aware navigation research.

Human-Aware NavigationVision-and-Language Navigation

Yifei Dong, Fengyi Wu, Qi He · 233 views · Mar 18, 2025

DONIC: Generalist Dexterous Bimanual Tracking Control from Human Demonstrations

Despite recent progress in dexterous reference tracking, existing models focus primarily on single-hand manipulation and a limited range of interactions, leaving the bimanual dexterous behaviors common in real life unaddressed. The scarcity of contact-rich bimanual manipulation data and the increased complexity of coordinating contacts make developing a generalist bimanual tracker particularly challenging. Therefore, we present 1) DONIC, the first generalist bimanual tracker jointly trained across operation categories and human-motion sources, together with 2) BiHOI, a multi-view dataset of 1,575 sequences spanning 229 objects and emphasizing sustained bimanual interaction, where reconstructed interactions are retargeted into robot references. DONIC combines proximal policy optimization with behavioral cloning in a feedback loop that recycles verified successful executions into training demonstrations. An object-centered per-hand reward maintains object-motion feedback while supervising each hand according to its reference contact role. Clip-Centered Advantages remove reference-dependent advantage offsets while preserving within-trajectory differences. Together, these designs enable one single policy to learn coordinated manipulation across heterogeneous references. Without finetuning, a single checkpoint achieves 86.79% success on unseen trajectories of training objects and 57.32% on held-out objects, surpassing the state-of-the-art bimanual tracking baselines by 77.36 and 45.73 percentage points.

Dexterous ManipulationBimanual Manipulation

Mingze Sun, Ziang Cao, Yuan Fei · 242 views · Oct 21, 2026

T-WLA: Bridging Slow Vision and Fast Touch

T-WLA shares a history-aware visual memory between a slow vision-language-action policy and a fast tactile residual editor. The visual branch produces an action chunk and memory at 39.7 Hz; the lightweight tactile editor reuses them to correct only unexecuted actions at 119.1 Hz. It improves tactile-conditioned hand-motion prediction and contact-rich manipulation, reaching 50.0% average success on four real-robot tasks with 30 demonstrations each.

Tactile SensingVisuotactile Manipulation

Kaichen Zhou, Yuzhen Chen, Hannah Yao · 310 views · Oct 20, 2026

Vision as Proprioception: Estimating Humanoid Robot Pose from Video

HumanoidPose combines frozen human-pose features with robot kinematics to recover full-body joint configurations and world-space motion from monocular video across embodiments.

HumanoidPose estimation

Jiaxi Jiang, Matthias Heyrman, René Zurbrügg · 261 views · Oct 18, 2026

3D Point World Models: Point Completion Enables More Accurate Dynamics Learning

Learning predictive models of the world enables robotic control through planning and can allow robots to improvise on new tasks. Video-based dynamics models lack explicit 3D structure and accumulate geometric inconsistencies, while existing partial-point-cloud world models remain sensitive to occlusion and rollout drift. 3DPWM first completes segmented object geometry from a single calibrated RGB-D view and robot joint angles, then learns action-conditioned per-point dynamics over the completed scene. Experiments across seven datasets and two embodiments show accurate 100-300 step rollouts, stronger reward correlation, improved open-loop and closed-loop planning, adaptation to an unseen long-horizon task, and sim-to-real transfer.

3D World ModelPoint Cloud Completion

Skand Peri, Hung Nguyen, Chanho Kim · 220 views · Jun 30, 2026

Thinker-WM: A Unified World Model with Shared Multimodal Representations for Embodied Intelligence

World models have emerged as a promising paradigm for embodied AI, yet existing systems often remain fragmented in the modalities they model and the functions they support. In particular, visual dynamics, geometric structure, and executable action are commonly coupled through separate interfaces or downstream modules rather than being jointly represented and evolved within a shared dynamics model. In this work, we present Thinker-WM, a unified multimodal world model that jointly models and generates RGB, metric depth, and robot action through a shared intermediate Unified Dynamics Transformer with modality-specific encoders and decoders. The model performs multimodal unification at the representation level: appearance, geometry, and action are transformed into a common dynamic representation and jointly evolved by shared parameters, while their distinct output structures are preserved. The same parameterization supports multiple functions, including joint video-action generation, world prediction, policy inference, action-conditioned generation, video-to-action generation, and inverse dynamics. To examine whether the shared representation captures action-related information, we introduce a lightweight motion probe that shows the frozen visual representation contains recoverable motion information. Thinker-WM is trained on a heterogeneous corpus comprising approximately 20K hours of egocentric video, 10K hours of real-robot data, and 1K hours of simulation data, including approximately 4K hours of real-robot data with aligned depth supervision. Extensive experiments demonstrate strong performance across embodied world-model evaluations. Thinker-WM ranks first on EWMBench, WorldOlympiad and WorldArena, while controlled studies further demonstrate the benefits of action conditioning, geometric interaction, and multi-function training within the shared representation.

World ModelsEmbodied AI

Thinker Team, Peng Gao, Jichao Jiao · 285 views · Oct 13, 2026

SuperMimic: Scaling Humanoid Whole-Body Control with One Differentiable World Model

SuperMimic fits one differentiable world model from simulator transitions and uses short imagined rollouts to train a general humanoid tracker. It surpasses SONIC in three hours on one H100, then reaches zero failures on 129,514 motions after 9.8 hours on two H100s.

HumanoidWorld Models

Qiayuan Liao, Yiyang Shao, Junfeng Long · 1.5k views · Oct 20, 2026

SimpleVLN: Towards Simple and Frontier Vision-Language Navigation

Vision-and-Language Navigation (VLN) requires embodied agents to ground language in visual observations and make long-horizon decisions. Existing methods often address these demands through task-specific architectures. As multimodal large language models (MLLMs) become increasingly capable, an important question remains: Have we truly unlocked the potential of native MLLMs for VLN before introducing task-specific architecture? We present SimpleVLN, an architecture-preserving framework developed through studies of navigation formulation, data construction, and optimization. At the formulation level, SimpleVLN combines single-turn decision inputs with fixed-budget global-local memory to efficiently use navigation history without unbounded context growth, while action chunks enable efficient output of multiple primitive actions per model invocation. For data construction, we combine expert demonstrations with policy-dependent DAgger experience and develop a scalable synthesis pipeline that generates large-scale tasks for reinforcement learning. At the optimization level, a progressive IL-DAgger-RL pipeline advances from action imitation to task-level outcome optimization. Comprehensive experiments validate the effectiveness of these three aspects. We instantiate SimpleVLN across multiple MLLM backbones and model sizes, demonstrating its generality. SimpleVLN achieves 76.7% and 80.1% success rates on the R2R-CE and RxR-CE splits, respectively, substantially outperforming existing methods. These results show that the navigation capabilities of native MLLMs can be effectively unlocked without task-specific architectural designs.

Vision-language navigationEmbodied AI

Lingyi Hong, Weiren Cai, Jiayu Chen · 316 views · Oct 16, 2026

As an Amazon Associate, we earn from qualifying purchases.