GLENS uses diffusion models on solver iterates to generate high-quality and diverse initial guesses for multimodal non-convex optimization, leading to faster solver convergence.
Canonical reference
Can vehicle motion planning generalize to realistic long-tail scenarios? In2024 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), pages 5388–5395, 2024
Canonical reference. 73% of citing Pith papers cite this work as background.
60
Pith papers citing it
262
external citations · external index
Background
73% of classified citations
citation-role summary
background 12
method 2
other 1
citation-polarity summary
representative citing papers
MemPoison enables stealthy memory poisoning in LLM agents via dialogue by using semantic relational bridges, entity masquerading, and joint embedding optimization to bypass selective extraction and rewriting, achieving up to 0.95 attack success rate.
MDrive benchmark shows multi-agent cooperative driving systems generally outperform single-agent ones in closed-loop settings but perception sharing does not always improve planning and negotiation can harm performance in complex traffic.
BehaviorBench reveals that self-play RL policies for autonomous driving overfit to their training traffic agents and do not generalize to other behaviors, motivating a hybrid rule-based plus learned planner.
Training scene flow networks on 940k synthetic CARLA LiDAR frames transfers zero-shot to real benchmarks and makes 5% real labels beat 20%.
The virtual object MPC framework enables stable shared teleoperation for transporting up to nine objects, cutting sliding distance by 72.45% and eliminating tip-overs compared to baseline.
A video foundation model trained on human demonstrations generates zero-shot plans that convert to executable robot actions on novel scenes and tasks.
Rule-VLN injects 177 regulatory signs into Touchdown-scale urban graphs; SNRM’s VLM perception plus mental-map detours cuts constraint violations ~19% and raises task completion ~6% zero-shot.
A closed-loop multi-agent LLM framework enables heterogeneous robots to collaboratively manipulate objects by decomposing tasks, grounding actions via visual tools, and recovering from execution failures hierarchically.
Zeta* and Zeta*-SIPP use elliptical expansion and improved visibility to deliver optimal any-angle paths on grids, with Zeta*-SIPP over 20x faster than prior dynamic planners.
X-Morph retargets human motions to kinematically plausible references for multiple legged morphologies, trains privileged RL trackers, and distills them into deployable policies that generalize and enable teleoperation and text-conditioned generation.
ReGuide is a self-improving framework that uses phase-conditioned guidance to generate corrective rollouts and absorbs successful ones back into diffusion policy training, yielding 1.3-7.7x success gains on Robomimic tasks.
The paper fine-tunes Qwen3.5-4B as a driving VLA using serialized decision traces from rule-based planners, reporting reduced ADE and miss rate on a simulator benchmark with camera inputs.
LAFM adapts the source distribution in flow matching policies via a latent action model to better match fragmented robotic action spaces, claiming 23.4% higher real-world success and 10.4% on LIBERO-90 while beating larger pre-trained models.
HilDA pre-trains LiDAR backbones via multi-layer and global distillation from vision models plus temporal occupancy diffusion, yielding SOTA results on detection, flow, and occupancy tasks.
A VLA policy using view-selective visual routing and interaction-aware action MoE improves average success by 27.7% in simulation and 43.3% in real-world bimanual tasks over monolithic baselines.
SpaceVLN proposes a stagewise closed-loop framework using Spatial Cognitive Memory and Spatial-CoT for zero-shot vision-and-language navigation and object-goal navigation, reporting SOTA results on R2R-CE, RxR-CE, GN-Bench, and HM3D-OVON plus real-robot tests.
RGB-S projects tactile contacts onto images as force-modulated Gaussian saliency maps via kinematics and zero-initialized conditioning, raising real-world occluded dexterous manipulation success by 26.7 percentage points over implicit baselines.
Perceptive BFM grounds human motion priors in robot terrain perception via terrain-conformal reference synthesis and teacher-student transfer from adapted to raw-reference tracking.
HORIZON is a recoverability-governed checkpointed frontier curriculum for on-policy physical-domain scaling on quadruped locomotion that identifies three regularities: uneven widening, non-monotonic composition, and the necessity of joint on-policy interaction.
AFUN predicts task-conditional functional masks and 3D post-contact motion curves from RGB-D and language, trained via a standardized multi-source data pipeline, and reports large gains over baselines on segmentation, contact prediction, and motion tasks.
POIROT protocol repurposes agents in LLM multi-agent systems as an internal diagnostic layer for failure detection, outperforming single-LLM evaluators with gains that increase with complexity, agent count, and fault types.
Per-Frame Deep Sets enables scaling single-sphere to five-sphere transport on a quadruped by performing permutation-invariant pooling within each history frame, reaching 100% no-drop success in simulation where standard encoders plateau.
A hierarchical multi-robot motion planner that refines workspace decompositions to enable scalable coordination through discrete search over smaller decoupled subproblems.
citing papers explorer
-
GLENS: Global Search via Learning from Solver Iterates with Diffusion Models
GLENS uses diffusion models on solver iterates to generate high-quality and diverse initial guesses for multimodal non-convex optimization, leading to faster solver convergence.
-
Hijacking Agent Memory: Stealthy Trojan Attacks Through Conversational Interaction
MemPoison enables stealthy memory poisoning in LLM agents via dialogue by using semantic relational bridges, entity masquerading, and joint embedding optimization to bypass selective extraction and rewriting, achieving up to 0.95 attack success rate.
-
MDrive: Benchmarking Closed-Loop Cooperative Driving for End-to-End Multi-agent Systems
MDrive benchmark shows multi-agent cooperative driving systems generally outperform single-agent ones in closed-loop settings but perception sharing does not always improve planning and negotiation can harm performance in complex traffic.
-
Beyond Self-Play and Scale: A Behavior Benchmark for Generalization in Autonomous Driving
BehaviorBench reveals that self-play RL policies for autonomous driving overfit to their training traffic agents and do not generalize to other behaviors, motivating a hybrid rule-based plus learned planner.
-
SynFlow: Scaling Up LiDAR Scene Flow Estimation with Synthetic Data
Training scene flow networks on 940k synthetic CARLA LiDAR frames transfers zero-shot to real benchmarks and makes 5% real labels beat 20%.
-
Towards Multi-Object Nonprehensile Transportation via Shared Teleoperation: A Framework Based on Virtual Object Model Predictive Control
The virtual object MPC framework enables stable shared teleoperation for transporting up to nine objects, cutting sliding distance by 72.45% and eliminating tip-overs compared to baseline.
-
Large Video Planner Enables Generalizable Robot Control
A video foundation model trained on human demonstrations generates zero-shot plans that convert to executable robot actions on novel scenes and tasks.
-
Rule-VLN: Bridging Perception and Compliance via Semantic Reasoning and Geometric Rectification
Rule-VLN injects 177 regulatory signs into Touchdown-scale urban graphs; SNRM’s VLM perception plus mental-map detours cuts constraint violations ~19% and raises task completion ~6% zero-shot.
-
A Closed-Loop Multi-Agent Framework for Robust Multi-Robot Manipulation
A closed-loop multi-agent LLM framework enables heterogeneous robots to collaboratively manipulate objects by decomposing tasks, grounding actions via visual tools, and recovering from execution failures hierarchically.
-
Optimal any-angle path planning in static and dynamic environments
Zeta* and Zeta*-SIPP use elliptical expansion and improved visibility to deliver optimal any-angle paths on grids, with Zeta*-SIPP over 20x faster than prior dynamic planners.
-
X-Morph: Human Motion Priors for Scalable Robot Learning Across Morphologies
X-Morph retargets human motions to kinematically plausible references for multiple legged morphologies, trains privileged RL trackers, and distills them into deployable policies that generalize and enable teleoperation and text-conditioned generation.
-
ReGuide: From Test-Time Guidance to Self-Improving Diffusion Policies
ReGuide is a self-improving framework that uses phase-conditioned guidance to generate corrective rollouts and absorbs successful ones back into diffusion policy training, yielding 1.3-7.7x success gains on Robomimic tasks.
-
Neuro-Symbolic Drive: Rule-Grounded Faithful Reasoning for Driving VLAs
The paper fine-tunes Qwen3.5-4B as a driving VLA using serialized decision traces from rule-based planners, reporting reduced ADE and miss rate on a simulator benchmark with camera inputs.
-
Flowing With Purpose: Latent Action Guided Flow Matching Policies For Robotic Manipulation
LAFM adapts the source distribution in flow matching policies via a latent action model to better match fragmented robotic action spaces, claiming 23.4% higher real-world success and 10.4% on LIBERO-90 while beating larger pre-trained models.
-
HilDA: Hierarchical Distillation with Diffusion for Advancing Self-Supervised LiDAR Pre-training
HilDA pre-trains LiDAR backbones via multi-layer and global distillation from vision models plus temporal occupancy diffusion, yielding SOTA results on detection, flow, and occupancy tasks.
-
See Selectively, Act Adaptively: Dual-Level Structural Decomposition for Bimanual Robot Manipulation
A VLA policy using view-selective visual routing and interaction-aware action MoE improves average success by 27.7% in simulation and 43.3% in real-world bimanual tasks over monolithic baselines.
-
SpaceVLN: A Zero-Shot Vision-and-Language Navigation Agent with Online Spatial Cognitive Memory and Reasoning
SpaceVLN proposes a stagewise closed-loop framework using Spatial Cognitive Memory and Spatial-CoT for zero-shot vision-and-language navigation and object-goal navigation, reporting SOTA results on R2R-CE, RxR-CE, GN-Bench, and HM3D-OVON plus real-robot tests.
-
RGB-S: Image-Aligned Tactile Saliency for Robust Dexterous Manipulation
RGB-S projects tactile contacts onto images as force-modulated Gaussian saliency maps via kinematics and zero-initialized conditioning, raising real-world occluded dexterous manipulation success by 26.7 percentage points over implicit baselines.
-
Perceptive Behavior Foundation Model: Adapting Human Motion Priors to Robot-Centric Terrain
Perceptive BFM grounds human motion priors in robot terrain perception via terrain-conformal reference synthesis and teacher-student transfer from adapted to raw-reference tracking.
-
HORIZON: Recoverability-Governed Curriculum for Physical-Domain Scaling
HORIZON is a recoverability-governed checkpointed frontier curriculum for on-policy physical-domain scaling on quadruped locomotion that identifies three regularities: uneven widening, non-monotonic composition, and the necessity of joint on-policy interaction.
-
AFUN: Towards an Affordance Foundation Model for Functionality Understanding
AFUN predicts task-conditional functional masks and 3D post-contact motion curves from RGB-D and language, trained via a standardized multi-source data pipeline, and reports large gains over baselines on segmentation, contact prediction, and motion tasks.
-
POIROT: Interrogating Agents for Failure Detection in Multi-Agent Systems
POIROT protocol repurposes agents in LLM multi-agent systems as an internal diagnostic layer for failure detection, outperforming single-LLM evaluators with gains that increase with complexity, agent count, and fault types.
-
S2M-Trek: From Single to Multi-Sphere Transport via Per-Frame Deep Sets on a Wheel-Legged Robot
Per-Frame Deep Sets enables scaling single-sphere to five-sphere transport on a quadruped by performing permutation-invariant pooling within each history frame, reaching 100% no-drop success in simulation where standard encoders plateau.
-
Scalable Multi-robot Motion Planning via Hierarchical Subproblem Expansion and Workspace Decomposition Refinement
A hierarchical multi-robot motion planner that refines workspace decompositions to enable scalable coordination through discrete search over smaller decoupled subproblems.
-
Bimodal Synchronization Performance: Why Noise and Sparse Connectivity Can Improve Collective Timing
In a discrete firefly-inspired oscillator model, synchronization is bimodal near a critical flash/quorum balance, and noise or sparser connectivity can rescue runs trapped in phase-shifted clusters.
-
Formal Methods Meet LLMs: Auditing, Monitoring, and Intervention for Compliance of Advanced AI Systems
Combines LTL formal methods with LLMs for auditing, predictive monitoring, and runtime intervention on temporally extended behavioral constraints, outperforming LLM baselines and reducing violations.
-
Robust Instruction Compliance in Cooperative Multi-Agent Reinforcement Learning
MAVIC corrects Bellman backups at instruction boundaries by adjusting the incoming objective and restoring continuation value, enabling consistent estimation under stochastic instruction switching in cooperative MARL.
-
Control of Fully Actuated Aerial Vehicles: A Comparison of Model-based and Sensor-based Dynamic Inversion
On a fully actuated hexarotor, sensor-based INDI outperforms model-based geometric NDI under mismatches, gusts, and sensor degradation with lower position errors, but NDI tracks attitude better at reduced control rates, providing the first experimental full-pose INDI validation with decoupled axes.
-
VISOR: A Vision-Language Model-based Test Oracle for Testing Robots
VISOR is a VLM-based automated test oracle that evaluates robot task correctness and quality from videos while reporting its own uncertainty, tested on GPT and Gemini across four tasks and over 1000 videos with Gemini showing higher recall and GPT higher precision but low uncertainty-correctness tie
-
High Precision Hydraulic Excavator Control for Heavy-Duty Grading
Autonomous excavator controller achieves 1.8 cm RMSE in heavy-duty grading across different hydraulic architectures, outperforming commercial solutions by a factor of 2.6 in precision while better utilizing machine pressure.
-
MAG-VLAQ: Multi-modal Aerial-Ground Query Aggregation for Cross-View Place Recognition
MAG-VLAQ fuses multi-modal ground and aerial data via ODE-conditioned vector-of-locally-aggregated-queries to nearly double recall@1 on aerial-ground place recognition benchmarks.
-
Dual-Timescale Memory in a Spiking Neuron-Astrocyte Network for Efficient Navigation
A neuron-astrocyte network with dual-timescale memory reduces median path lengths up to sixfold in partially observable grid-world navigation tasks.
-
ViserDex: Visual Sim-to-Real for Robust Dexterous In-hand Reorientation
A framework using 3D Gaussian Splatting for visual domain randomization enables robust monocular RGB-based dexterous in-hand reorientation on real hardware for multiple objects under varied lighting.
-
WARPED: Wrist-Aligned Rendering for Robot Policy Learning from Egocentric Human Demonstrations
WARPED synthesizes realistic wrist-view observations from monocular egocentric human videos via foundation models, hand-object tracking, retargeting, and Gaussian Splatting to train visuomotor policies that match teleoperation success rates on five tabletop tasks with 5-8x less collection effort.
-
House of Dextra: Cross-embodied Co-design for Dexterous Hands
A cross-embodied co-design framework learns task-specific hand morphologies and control policies, achieving sim-to-real rotation up to 3.3 rad/s and full fabrication in under 24 hours.
-
RAPTOR: A Foundation Policy for Quadrotor Control
A 2084-parameter recurrent policy trained by distilling 1000 RL teacher policies enables zero-shot control across 10 real quadrotors differing in mass, motors, frames, propellers, and flight controllers.
-
ViTacFormer: Learning Cross-Modal Representation for Visuo-Tactile Dexterous Manipulation
ViTacFormer learns a cross-modal visuo-tactile latent space with autoregressive tactile prediction and an easy-to-hard curriculum, then uses the representation for imitation learning that yields ~50% higher success and the first reported 11-stage, 2.5-minute autonomous dexterous tasks.
-
AutoVLA: A Vision-Language-Action Model for End-to-End Autonomous Driving with Adaptive Reasoning and Reinforcement Fine-Tuning
AutoVLA unifies semantic reasoning and trajectory planning in one autoregressive VLA model for end-to-end autonomous driving by tokenizing trajectories into discrete actions and using GRPO reinforcement fine-tuning to adaptively reduce unnecessary reasoning.
-
CSAR: Containerized System Architecture for Robotics
CSAR organizes robotics lab infrastructure into three layers — a stable core, an LXC/LXD multi-user platform, and disposable compute containers — and demonstrates it on edge-offloaded SLAM and semantic mapping.
-
RefGlass-GS: A UAV-Enabled Fusion Framework for Photorealistic, Semantic and Interactive Digitization of Reflective Glass Facades via Gaussian Splatting
RefGlass-GS is a fusion framework using UAV data, MAP-based panel segmentation, viewpoint optimization, and modified Gaussian Splatting with Reflection MLP to achieve improved photorealistic and semantic modeling of reflective glass facades.
-
Any-Body Guard: Universal Safeguarding for Manipulation Policies via Action Masking
X-Safe masks actions in configuration space using forward kinematics and quasi-static object models to give probabilistic collision-avoidance guarantees that transfer across robot embodiments without per-setup engineering.
-
RoboLineage: Agent-Native Data Lifecycle Governance Across Robot Policy Iterations
RoboLineage introduces an agent-native data lifecycle governance system that represents robot policy iteration steps as typed lineage artifacts to improve speed and auditability in real-robot workflows.
-
Online Learning of Robust Legged Odometry with Minimal Exteroceptive Supervision
Develops an online-learned proprioceptive velocity model supervised by exteroceptive pipelines, fused via InEKF with IMU for resilient quadruped odometry that adapts without explicit calibration.
-
Sparse2Act: Learning Action-Aligned Sparse 3D Representations for Cross-Domain Robot Manipulation
Sparse2Act pretrains sparse 3D encoders via masked action-alignment supervision, yielding reusable representations that reach 86.9% success on LIBERO-10 and enable cross-domain transfer.
-
LLM-ALSO: LLM-Driven Adaptive Learning-Signal Optimization for Multi-Agent Reinforcement Learning
LLM-ALSO is an iterative LLM-driven framework for adaptive reward shaping in cooperative MARL that diagnoses failures from sparse metrics, proposes configurations, and validates them via short-horizon branches before applying updates.
-
KAPPS: A knowledge-based CPPS Architecture for the Circular Factory
KAPPS is a knowledge-based CPPS architecture that uses an ontology-grounded knowledge graph as the unifying data backbone and authoritative write-time state for handling uncertainty in circular manufacturing, demonstrated via anomaly detection and constraint enforcement use cases.
-
Introducing Environmental Constraints to Grasping Strategies for Paper-Like Flexible Materials Using a Soft Gripper
Systematic grasping strategies for paper-like materials are developed and tested with a soft gripper by exploiting environmental constraints to improve force control and success rates.
-
AgentRx: A Benchmark Study of LLM Agents for Multimodal Clinical Prediction Tasks
Single-agent LLM frameworks outperform naive multi-agent systems in multimodal clinical risk prediction tasks and are better calibrated.
-
What Will Happen Next: Large Models-Driven Deduction for Emergency Instances
WLDS applies large models with factual and logical calibration to produce diverse text-and-image deductions of emergency scenarios beyond what traditional fixed simulations can generate.
-
Evaluating Generative Models as Interactive Emergent Representations of Human-Like Collaborative Behavior
Embodied LLM agents exhibit emergent collaborative behaviors indicating mental models of partners in a color-matching game, detected via LLM judges and supported by positive user feedback.