{
 "generated": "2026-11-07",
 "count": 14,
 "note": "Every value carries a verbatim quote verified against the paper text. Missing values are not reported by the paper.",
 "rows": [
  {
   "task": "human pose estimation from mmWave radar",
   "robot": null,
   "sim_platform": "4DPS",
   "transfer": "zero-shot",
   "sim_result": null,
   "real_result": {
    "value": "21.9 cm",
    "metric": "MPJPE on real RT-Pose data for a pose model trained on rendered data alone",
    "quote": "On RT-Pose, a pose model trained on them alone reaches 21.9 cm MPJPE on real data, against 16.7 cm for real training and 127.6 cm for the RF-Genesis baseline",
    "kind": "other",
    "low": null,
    "high": null
   },
   "missing": [
    "rendered-data test MPJPE for the rendered-trained pose model",
    "absolute AP for real-trained and rendered-trained HuPR pose models"
   ],
   "checkable": true,
   "delta_pp": null,
   "arxiv": "2611.08342",
   "source": "https://arxiv.org/html/2611.08342",
   "verified": true,
   "extracted_at": "2026-11-07T00:40:44+00:00",
   "title": "4D Point Splatting for mmWave Radar Micro-Doppler Differentiable Rendering",
   "authors": [
    "Adnan Armouti",
    "Guanyue Qian",
    "Rajalakshmi Nandakumar"
   ],
   "published": "2026-11-05"
  },
  {
   "task": "dual-arm object following and dexterous in-hand manipulation",
   "robot": null,
   "sim_platform": "IsaacLab",
   "transfer": "not stated",
   "sim_result": null,
   "real_result": null,
   "missing": [
    "success rate in simulation",
    "success rate on real robot",
    "sim-to-real success-rate gap"
   ],
   "checkable": true,
   "delta_pp": null,
   "arxiv": "2611.07881",
   "source": "https://arxiv.org/html/2611.07881",
   "verified": true,
   "extracted_at": "2026-11-07T00:41:19+00:00",
   "title": "TactileLab: Sim-to-Real Dexterous Robot Manipulation with Efficient Shear-Sensitive Tactile Simulation",
   "authors": [
    "Yijiong Lin",
    "Wenhao He",
    "Toby Jack",
    "Max Yang",
    "Haoran Li",
    "Nathan F. Lepora"
   ],
   "published": "2026-11-05"
  },
  {
   "task": "Keypoint positioning of deformable linear objects",
   "robot": null,
   "sim_platform": "MuJoCo Playground",
   "transfer": "zero-shot",
   "sim_result": {
    "value": "96.4% for single-point tracking, 98.0% for disturbance recovery, and 88.8% for two-point tracking",
    "metric": "simulation success rate",
    "quote": "Policies trained with observation noise, delay, domain randomization, and disturbances achieve simulation success rates of 96.4% for single-point tracking, 98.0% for disturbance recovery, and 88.8% for two-point tracking.",
    "kind": "range",
    "low": 96.4,
    "high": 98.0
   },
   "real_result": {
    "value": "9/10, 10/10, and 8/10 physical trials",
    "metric": "real-world success rate",
    "quote": "Using markerless RGB-D keypoint estimation, the policies transfer without real-world fine-tuning, succeeding in 9/10, 10/10, and 8/10 physical trials, respectively.",
    "kind": "other",
    "low": null,
    "high": null
   },
   "missing": [
    "real-world completion time",
    "perception keypoint error",
    "sim-to-real success-rate gap"
   ],
   "checkable": true,
   "delta_pp": null,
   "arxiv": "2611.07261",
   "source": "https://arxiv.org/html/2611.07261",
   "verified": true,
   "extracted_at": "2026-11-07T00:41:50+00:00",
   "title": "An Empirical Study of Reinforcement Learning for Keypoint Positioning of Deformable Linear Objects",
   "authors": [
    "Vinay Rodrigues",
    "Sara Zeynalpour",
    "Gokhan Solak",
    "Sebastian Hjorth",
    "Arash Ajoudani"
   ],
   "published": "2026-11-05"
  },
  {
   "task": "autonomous mobile robot navigation in human environments",
   "robot": "TIAGo",
   "sim_platform": "PeopleFlow (Gazebo-based social simulator)",
   "transfer": "fine-tuned",
   "sim_result": null,
   "real_result": null,
   "missing": [
    "quantitative simulation task success rate",
    "quantitative real-world task success rate",
    "quantitative proxemic compliance rate"
   ],
   "checkable": false,
   "delta_pp": null,
   "arxiv": "2611.04027",
   "source": "https://arxiv.org/html/2611.04027",
   "verified": true,
   "extracted_at": "2026-11-07T00:42:52+00:00",
   "title": "A Practical Causal Framework for Autonomous Mobile Robots: From Simulation to Real-World Human Environments",
   "authors": [
    "Luca Castri",
    "Gloria Beraldo",
    "Nicola Bellotto"
   ],
   "published": "2026-11-02"
  },
  {
   "task": "Sim-to-real V2X LiDAR object detection",
   "robot": "Ouster OS1-64",
   "sim_platform": "UrbanTwin",
   "transfer": "zero-shot",
   "sim_result": {
    "value": "0.9251",
    "metric": "realism score",
    "quote": "Our final entry scores 0.7457 combined, from a detection mAP of 0.4382 and a realism score of 0.9251.",
    "kind": "other",
    "low": null,
    "high": null
   },
   "real_result": {
    "value": "0.4382",
    "metric": "detection mAP",
    "quote": "Our final entry scores 0.7457 combined, from a detection mAP of 0.4382 and a realism score of 0.9251.",
    "kind": "other",
    "low": null,
    "high": null
   },
   "missing": [
    "synthetic detection mAP",
    "final per-class AP on real for vehicle, pedestrian, truck"
   ],
   "checkable": true,
   "delta_pp": null,
   "arxiv": "2611.03940",
   "source": "https://arxiv.org/html/2611.03940",
   "verified": true,
   "extracted_at": "2026-11-07T00:44:07+00:00",
   "title": "Coordinate Frame Recovery, Per Class NMS and Null Return Modelling for Sim2Real V2X LiDAR Detection",
   "authors": [
    "Aman Upganlawar"
   ],
   "published": "2026-11-02"
  },
  {
   "task": "Counterfactual spatial imagination for navigation and manipulation",
   "robot": null,
   "sim_platform": "AI2-THOR and ProcTHOR",
   "transfer": "fine-tuned",
   "sim_result": {
    "value": "29.5 percentage points",
    "metric": "Accuracy improvement of Qwen3-VL-8B after fine-tuning on FlowBench-Sim-Train, evaluated in simulation",
    "quote": "fine-tuning Qwen3-VL-8B on FlowBench-Sim-Train raises its accuracy by 29.5 percentage points in simulation and 16.9 on real images, a strong sim-to-real transfer.",
    "kind": "other",
    "low": null,
    "high": null
   },
   "real_result": {
    "value": "16.9 on real images",
    "metric": "Accuracy improvement of Qwen3-VL-8B after fine-tuning on FlowBench-Sim-Train, evaluated on FlowBench-Real-Test",
    "quote": "fine-tuning Qwen3-VL-8B on FlowBench-Sim-Train raises its accuracy by 29.5 percentage points in simulation and 16.9 on real images, a strong sim-to-real transfer.",
    "kind": "other",
    "low": null,
    "high": null
   },
   "missing": [
    "Absolute post-fine-tune accuracy of Qwen3-VL-8B on FlowBench-Sim-Test",
    "Absolute post-fine-tune accuracy of Qwen3-VL-8B on FlowBench-Real-Test",
    "Pre-fine-tune accuracy of Qwen3-VL-8B on FlowBench-Real-Test"
   ],
   "checkable": true,
   "delta_pp": null,
   "arxiv": "2611.03937",
   "source": "https://arxiv.org/html/2611.03937",
   "verified": true,
   "extracted_at": "2026-11-07T00:45:13+00:00",
   "title": "FlowBench: Counterfactual Imagination Training for Spatial Intelligence",
   "authors": [
    "Yifan Liu",
    "Baihan Zhang",
    "Khai Loong Aw",
    "Klemen Kotar",
    "Daniel L. K. Yamins"
   ],
   "published": "2026-11-02"
  },
  {
   "task": "Sim2Real roadside LiDAR detection",
   "robot": null,
   "sim_platform": null,
   "transfer": "zero-shot",
   "sim_result": {
    "value": "0.9126",
    "metric": "realism score",
    "quote": "Our final entry scores 0.4793 combined, from a detection mAP of 0.1333 and a realism score of 0.9126.",
    "kind": "other",
    "low": null,
    "high": null
   },
   "real_result": {
    "value": "0.1333",
    "metric": "detection mAP",
    "quote": "Our final entry scores 0.4793 combined, from a detection mAP of 0.1333 and a realism score of 0.9126.",
    "kind": "other",
    "low": null,
    "high": null
   },
   "missing": [
    "detection mAP on synthetic validation data",
    "per-class AP for all seven classes on hidden real frames",
    "individual Chamfer, MMD, EMD, and FPD scores for the submitted synthetic point clouds"
   ],
   "checkable": true,
   "delta_pp": null,
   "arxiv": "2611.03936",
   "source": "https://arxiv.org/html/2611.03936",
   "verified": true,
   "extracted_at": "2026-11-07T00:46:27+00:00",
   "title": "Geometric Refinement and Distributional Alignment for Sim2Real Roadside LiDAR Detection",
   "authors": [
    "Aman Upganlawar"
   ],
   "published": "2026-11-02"
  },
  {
   "task": "Endovascular navigation to right carotid targets",
   "robot": "custom-built endovascular robot",
   "sim_platform": "stEVE framework and SOFA",
   "transfer": "zero-shot",
   "sim_result": {
    "value": "81%",
    "metric": "success rate",
    "quote": "While fully autonomous navigation achieved 81% success in simulation, performance dropped to 23% in vitro .",
    "kind": "success_rate",
    "low": 81.0,
    "high": 81.0
   },
   "real_result": {
    "value": "23%",
    "metric": "success rate",
    "quote": "While fully autonomous navigation achieved 81% success in simulation, performance dropped to 23% in vitro .",
    "kind": "success_rate",
    "low": 23.0,
    "high": 23.0
   },
   "missing": [
    "normalized path ratio for simulation and in-vitro autonomous navigation",
    "procedure time for simulation and in-vitro autonomous navigation",
    "per-target success rates RICA versus RECA"
   ],
   "checkable": true,
   "delta_pp": 58.0,
   "arxiv": "2611.03446",
   "source": "https://arxiv.org/html/2611.03446",
   "verified": true,
   "extracted_at": "2026-11-07T00:46:47+00:00",
   "title": "Clinician-AI Collaboration in Endovascular Neurointerventions: In-Vitro User Study of Copilot Assistance Modes for Robotic Navigation Across the Autonomy Spectrum",
   "authors": [
    "Nikola Fischer",
    "Harry Robertshaw",
    "S. M. Hadi Sadati",
    "Xingyu Chen",
    "Weiyuan Deng",
    "Benjamin Jackson",
    "Michael Adofo Kwakye",
    "Sara Sciacca",
    "Naga Kandasamy",
    "Parthiban Balasundaram",
    "Sahan Guruge",
    "Charles Cummins",
    "Joe Leyon",
    "Nasr Abdelsalam",
    "Matthew Benger",
    "Andrew Falzon",
    "Alejandro Granados",
    "Christos Bergeles",
    "Thomas Booth"
   ],
   "published": "2026-11-02"
  },
  {
   "task": "multi-arm contact-rich manipulation",
   "robot": null,
   "sim_platform": null,
   "transfer": "zero-shot",
   "sim_result": {
    "value": "65.5%",
    "metric": "mean policy success",
    "quote": "We evaluate ScheduleGen on 28 simulation task variants, including seven externally authored RoboDojo tasks and settings inaccessible to prior systems; it achieves 79.6% mean data-generation success and trains policies with 65.5% mean success.",
    "kind": "success_rate",
    "low": 65.5,
    "high": 65.5
   },
   "real_result": {
    "value": "50–80%",
    "metric": "success across eight real-world variants",
    "quote": "ScheduleGen policies outperform those trained on a state-of-the-art data generation baseline, and achieve 50–80% success across eight real-world variants through zero-shot sim-to-real and real-to-sim-to-real transfer.",
    "kind": "range",
    "low": 50.0,
    "high": 80.0
   },
   "missing": [
    "real-world mean success rate",
    "per-task real-world success rates",
    "number of real-world trials per variant"
   ],
   "delta_pp": null,
   "arxiv": "2610.39429",
   "verified": true,
   "title": "ScheduleGen: Demonstration Generation via Skill Scheduling for Multi-Arm Imitation Learning",
   "authors": [
    "Neel Jawale",
    "Stephan Pleines",
    "Soha Pouya",
    "Yuke Zhu",
    "Ajay Mandlekar",
    "Caelan Garrett"
   ],
   "published": "2026-10-29"
  },
  {
   "task": "steep deformable slope traversal",
   "robot": "ERNEST rover",
   "sim_platform": "DARTS",
   "transfer": "online adaptation",
   "sim_result": {
    "value": "approximately 90%",
    "metric": "slope-climbing success",
    "quote": "In simulation, TERRA reaches approximately 90% slope-climbing success within five trials under severe out-of-distribution pretraining conditions, outperforming exploration baselines in reward, completion time, and energy to first success.",
    "kind": "success_rate",
    "low": 90.0,
    "high": 90.0
   },
   "real_result": {
    "value": "within three to five real-world trials",
    "metric": "recovery of coordinated active-suspension locomotion",
    "quote": "On the ERNEST rover, TERRA recovers coordinated active-suspension locomotion on a deformable Mars-simulant slope within three to five real-world trials.",
    "kind": "other",
    "low": null,
    "high": null
   },
   "missing": [
    "real-world slope-climbing success rate",
    "sim-to-real performance gap",
    "real-world energy or completion time"
   ],
   "delta_pp": null,
   "arxiv": "2610.38657",
   "verified": true,
   "title": "Explore Where It Matters: Targeted Exploration for Efficient Rover Adaptation",
   "authors": [
    "Jacob Levy",
    "Arthur Bouton",
    "Tristan D. Hassler",
    "Hari Nayar",
    "David Fridovich-Keil"
   ],
   "published": "2026-10-28"
  },
  {
   "task": "retractor take-over in open surgery",
   "robot": "Trossen WidowX AI arm",
   "sim_platform": "Isaac Sim",
   "transfer": "zero-shot",
   "sim_result": {
    "value": "94%",
    "metric": "grasp success rate on held-out simulated presentations",
    "quote": "The same checkpoint grasps 94 % of held-out simulated presentations, with a post-release tip offset of 8.0 ± \\pm 6.0 mm and axial twist of 3.8 ± \\pm 6.4 ∘ over grasped trials.",
    "kind": "success_rate",
    "low": 94.0,
    "high": 94.0
   },
   "real_result": {
    "value": "90%",
    "metric": "grasp success rate on surgeon-defined real presentations",
    "quote": "Deployed zero-shot, it grasps the retractor in 90 % of 50 surgeon-defined phantom presentations and preserves retraction at handoff in 78 %.",
    "kind": "success_rate",
    "low": 90.0,
    "high": 90.0
   },
   "missing": [
    "real post-release instrument tip offset",
    "real axial twist",
    "real-world instrument segmentation/plane-estimation error"
   ],
   "delta_pp": 4.0,
   "arxiv": "2610.38234",
   "verified": true,
   "title": "SimRetract: Zero-Shot Sim-to-Real Robotic Assistance for Retractor Take-Over in Open Surgery",
   "authors": [
    "Xucheng Wang",
    "Zhizhou Yang",
    "Sung Eun Kim",
    "Romain Hardy",
    "Michael P. Sampson",
    "Aakash Reddy",
    "Ankit Pal",
    "Xiaoman Zhang",
    "Luyang Luo",
    "Pranav Rajpurkar"
   ],
   "published": "2026-10-28"
  },
  {
   "task": "Image-goal navigation for drones with omnidirectional vision",
   "robot": "drone with Jetson Orin NX, OAK-4P-New quad-fisheye rig, Livox Mid-360s LiDAR",
   "sim_platform": "Habitat, IsaacSim",
   "transfer": "zero-shot",
   "sim_result": null,
   "real_result": null,
   "missing": [
    "simulation Success Rate (SR) on Habitat benchmarks",
    "real-world Success Rate (SR)",
    "collision rate in real-world flights"
   ],
   "delta_pp": null,
   "arxiv": "2610.36968",
   "verified": true,
   "title": "Omni-IGN: Onboard Image-Goal Navigation for Drones with Omnidirectional Vision",
   "authors": [
    "Zichen Yan",
    "Rui Huang",
    "Shenao Wang",
    "Yucong Wang",
    "Nuthasith Gerdpratoom",
    "Lin Zhao"
   ],
   "published": "2026-10-27"
  },
  {
   "task": "pressing buttons and moving sliders",
   "robot": "Unitree G1",
   "sim_platform": "MuJoCo",
   "transfer": "not stated",
   "sim_result": null,
   "real_result": null,
   "missing": [
    "manipulation success rate values for Midi Slider/Toggle Switch tasks",
    "sim-to-real transfer success rate"
   ],
   "delta_pp": null,
   "arxiv": "2610.36859",
   "verified": true,
   "title": "ActiveAlign: Aligning Egocentric Manipulation via Active Perception",
   "authors": [
    "Ri-Zhao Qiu",
    "Bowen Wen",
    "Yifeng Jiang",
    "Chen Tessler",
    "Eric Heiden",
    "Xue Bin Peng",
    "Xiaolong Wang",
    "Stan Birchfield",
    "Erwin Coumans"
   ],
   "published": "2026-10-27"
  },
  {
   "task": "soft robot shape reconstruction",
   "robot": "SoftSnap",
   "sim_platform": null,
   "transfer": "zero-shot",
   "sim_result": {
    "value": "93%",
    "metric": "shape reconstruction success rate",
    "quote": "On 5,000 test samples generated from previously unseen threading patterns, the method achieves a 93% shape reconstruction success rate under surrogate-based evaluation.",
    "kind": "success_rate",
    "low": 93.0,
    "high": 93.0
   },
   "real_result": null,
   "missing": [
    "physical reconstruction error",
    "real-world success rate",
    "sim-to-real error"
   ],
   "delta_pp": null,
   "arxiv": "2610.36681",
   "verified": true,
   "title": "From Biological Curves to Soft Modular Robots: Learning-Based Inverse Design with Uncertainty-Guided Refinement",
   "authors": [
    "Dixiang Wang",
    "Minghao Zha",
    "Arifur Rahman",
    "Yinhong Qin",
    "Yitao Jiang",
    "Shiyang Lu",
    "Zebing Mao",
    "Muhao Chen",
    "Luyang Zhao"
   ],
   "published": "2026-10-27"
  }
 ]
}