Pith. sign in

REVIEW 5 cited by

Self-Supervised Policy Adaptation during Deployment

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

arxiv 2007.04309 v3 pith:GYIVOERJ submitted 2020-07-08 cs.LG cs.CVcs.ROstat.ML

classification cs.LGcs.CVcs.ROstat.ML
keywords environmentenvironmentsdeploymentpolicyacrosschangesdifferentgeneralization
verification ladder T0 review T1 audit T2 compute T3 formal
0 comments
read the original abstract

In most real world scenarios, a policy trained by reinforcement learning in one environment needs to be deployed in another, potentially quite different environment. However, generalization across different environments is known to be hard. A natural solution would be to keep training after deployment in the new environment, but this cannot be done if the new environment offers no reward signal. Our work explores the use of self-supervision to allow the policy to continue training after deployment without using any rewards. While previous methods explicitly anticipate changes in the new environment, we assume no prior knowledge of those changes yet still obtain significant improvements. Empirical evaluations are performed on diverse simulation environments from DeepMind Control suite and ViZDoom, as well as real robotic manipulation tasks in continuously changing environments, taking observations from an uncalibrated camera. Our method improves generalization in 31 out of 36 environments across various tasks and outperforms domain randomization on a majority of environments.

Discussion (0). Sign in to comment.

Forward citations

Cited by 5 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. Forget, Anticipate and Adapt: Test Time Training for Long Videos

    cs.CV 2026-06 unverdicted novelty 7.0 of 10

    FFN performs TTT on multi-hour videos by restricting updates to three frames and using a surprise metric for adaptive window sizing, plus a new EpicTours dataset.

  2. Pix2Act: Image-Space Manipulation Policies with Equivariant Augmentation

    cs.RO 2026-07 conditional novelty 6.0 of 10

    Continuous multi-view image-space keypoint trajectories plus per-camera equivariant augmentation beat strong 3D and image baselines on MimicGen and real UR5 tasks.

  3. Test3R: Learning to Reconstruct 3D at Test Time

    cs.CV 2025-06 conditional novelty 6.0 of 10

    Test3R improves 3D reconstruction by optimizing visual prompts at test time so that pointmaps from different image pairs are geometrically consistent.

  4. Box Pose and Shape Estimation and Domain Adaptation for Large-Scale Warehouse Automation

    cs.RO 2025-07 conditional novelty 5.0 of 10

    BOSS uses certificate-checked pseudo-labels to self-train a stereo keypoint network, improving box pose and shape estimates on real warehouse data without manual labels.

  5. Steering Robots with Inference-Time Interactions

    cs.RO 2025-06 conditional novelty 4.0 of 10

    Frozen imitation policies can be steered at inference time via user interactions, with a diffusion-sampling method and a constraint-enforcing framework that provides formal task guarantees.

Pith tools