REVIEW 16 cited by
PoliFormer: Scaling On-Policy RL with Transformers Results in Masterful Navigators
Not yet reviewed by Pith; the record is open.
This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.
SPECIMEN: schema-true, not a live event
T0 review · schema-true
One-sentence machine reading of the paper's core claim.
pith:XXXXXXXX · record.json · timestamp
PoliFormer: Scaling On-Policy RL with Transformers Results in Masterful Navigators
read the original abstract
We present PoliFormer (Policy Transformer), an RGB-only indoor navigation agent trained end-to-end with reinforcement learning at scale that generalizes to the real-world without adaptation despite being trained purely in simulation. PoliFormer uses a foundational vision transformer encoder with a causal transformer decoder enabling long-term memory and reasoning. It is trained for hundreds of millions of interactions across diverse environments, leveraging parallelized, multi-machine rollouts for efficient training with high throughput. PoliFormer is a masterful navigator, producing state-of-the-art results across two distinct embodiments, the LoCoBot and Stretch RE-1 robots, and four navigation benchmarks. It breaks through the plateaus of previous work, achieving an unprecedented 85.5% success rate in object goal navigation on the CHORES-S benchmark, a 28.5% absolute improvement. PoliFormer can also be trivially extended to a variety of downstream applications such as object tracking, multi-object navigation, and open-vocabulary navigation with no finetuning.
Forward citations
Cited by 16 Pith papers
-
Where to Look: Can Foundation Models Reach a Target Viewpoint Through Active Exploration?
Introduces the TVR active viewpoint-matching task and TVRBench indoor simulation benchmark, where foundation models start at low single-digit success rates but reach 51.4% after visual-action SFT and multi-turn GRPO p...
-
Beyond Isolation: A Unified Benchmark for General-Purpose Navigation
OmniNavBench is a unified benchmark for general-purpose navigation featuring composite multi-skill instructions, support for humanoid, quadrupedal and wheeled robots, and 1779 human teleoperated trajectories across 17...
-
LangMap: A Human-Verified Benchmark for Hierarchical Open-Vocabulary Goal Navigation
LangMap is a human-verified navigation benchmark with 18K tasks spanning scene-, room-, region-, and instance-level goals in real-world 3D scans, covering 414 object categories.
-
ABot-N1: Toward a General Visual Language Navigation Foundation Model
ABot-N1 decouples navigation into a slow CoT-plus-pixel-goal reasoner and a fast waypoint controller, claiming state-of-the-art results on five VLN benchmarks and releasing two new urban navigation benchmarks.
-
ABot-N1: Toward a General Visual Language Navigation Foundation Model
A slow–fast VLN model that routes five navigation tasks through CoT plus image-space pixel goals reaches SOTA on established and new urban benchmarks, including 77.3% POI arrival.
-
ABot-N1: Toward a General Visual Language Navigation Foundation Model
ABot-N1 decouples VLN into a slow CoT reasoner that outputs pixel goals and a fast action expert, claiming large SOTA gains on urban POI and multi-task navigation.
-
SurveilNav: Collaborative Object Goal Navigation with Robot and Surveillance System
SurveilNav integrates robot local perception with multi-view surveillance for improved collaborative object goal navigation and reports SOTA results on HM3D.
-
NavOL: Navigation Policy with Online Imitation Learning
NavOL collects expert trajectory labels online from a global planner during policy rollouts in simulation to train a diffusion navigation policy, mitigating distribution shift and improving performance on visual navig...
-
Plan in Sandbox, Navigate in Open Worlds: Learning Physics-Grounded Abstracted Experience for Embodied Navigation
SAGE trains agents in physics-grounded semantic abstractions via RL with asymmetric clipping, achieving 53.21% LLM-Match Success on A-EQA (+9.7% over baseline) and encouraging physical robot transfer.
-
VLA-RL: Towards Masterful and General Robotic Manipulation with Scalable Reinforcement Learning
VLA-RL applies online RL to pretrained VLAs, yielding a 4.5% gain over strong baselines on 40 LIBERO manipulation tasks and matching commercial models like π₀-FAST.
-
Uni-NaVid: A Video-based Vision-Language-Action Model for Unifying Embodied Navigation Tasks
Uni-NaVid unifies diverse embodied navigation tasks into one video-based vision-language-action model trained on 3.6 million samples from four sub-tasks, achieving state-of-the-art performance on benchmarks and real-w...
-
Qwen-RobotNav Technical Report: A Scalable Navigation Model Designed for an Agentic Navigation System
Qwen-RobotNav introduces a parameterized navigation model supporting multiple task modes and controllable observation parameters, trained on 15.6M samples with vision-language co-training to achieve SOTA results on be...
-
Qwen-RobotNav Technical Report: A Scalable Navigation Model Designed for an Agentic Navigation System
Qwen-RobotNav provides a parameterized navigation model trained on 15.6M samples with vision-language co-training that achieves SOTA results on benchmarks and zero-shot transfer to real robots.
-
AllDayNav: Lifelong Navigation via Real-World Reinforcement Learning
AllDayNav encodes scene dynamics into a large model's parameters via RL and a multimodal memory, achieving near-100% success rates in lifelong navigation and outperforming map-based and VLM baselines.
-
Learning Category-level Last-meter Navigation from RGB Demonstrations of a Single-instance
An RGB-only imitation-learning policy trained on one chair generalizes to unseen chairs and environments for last-meter base positioning, though reported success depends on an added heuristic stopping rule and a 0.3 m...
-
SafeVLA: Towards Safety Alignment of Vision-Language-Action Model via Constrained Learning
SafeVLA applies constrained reinforcement learning via CMDP min-max optimization to VLAs, cutting safety violation costs by 83.58% while preserving task success on long-horizon mobile manipulation tasks.
discussion (0)
Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.