HarnessVLN: training-free embodied navigation agent sets SOTA on four benchmarks
Yang Chen · hf · 2026-09-16
- Approach: HarnessVLN is a zero-shot, training-free embodied navigation framework whose Agent Harness coordinates perception, retrieval, grounding, navigation, recovery, and termination via a unified tool interface.
- Verification: Planner proposals are validated against spatial evidence, geometric feasibility, and subgoal consistency; hierarchical event memory tracks progress while a persistent spatiotemporal graph stores reusable spatial evidence and failure annotations.
- Results: 60.8% / 53.9% / 76.0% / 59.3% success rates on R2R, RxR, HM3D-v2, and HM3D-OVON, surpassing prior training-free SOTA.
- Deployment: A replaceable Navigation Executor supports both instruction-following and object-goal navigation, demonstrated on a humanoid robot in real-world environments.
More from Embodied
- Wearing Whoop and Apple Watch on a single dual-slot band — tinyfool · 2026-11-02
- REA: async brain-cerebellum runtime for robots where policies never wait on the LLM — Pengfei Zhou · 2026-11-02
- Robot Combat Sports Coming to Japan with a Permanent Arena in Tokyo — cixliv · 2026-11-02
- REK robot fighting in the cage: humanoid bot in boxing gloves draws a crowd — broodsugar · 2026-11-02
- iPhone Duo: a folding iPhone mod impresses, but its hinge raises crackgate concerns — rudrank · 2026-11-02
- InflatableBots: University of Calgary's inflatable robots bring dynamic haptics to large-scale VR — lukas_m_ziegler · 2026-11-02