RynnValue: Scaling Robotic Value Foundation Models with Temporal Distance

Hugging Face Daily Papers Papers
robotics foundation-model temporal-distance reward-model manipulation open-source

Summary

RynnValue introduces an open-source robotic value foundation model that uses temporal distance as a scalable supervision target for reward learning, surpassing preference-supervised state-of-the-art on RBM-EVAL-OOD and improving real-world policy success rates.

General-purpose reward models are increasingly the bottleneck for scaling robot learning, yet the recipe for learning value-related capabilities from large-scale heterogeneous corpora remains underexplored. Existing approaches tie supervision to task-internal anchors such as preferences or normalized progress, none of which transfer cleanly across embodiments and data sources. We introduce RynnValue, an open-source value foundation model for robotic manipulation that replaces these anchors with temporal distance, the directed cost-to-go from an observation to the language-specified goal. Because temporal-distance labels can be derived directly from timestamps, RynnValue scales to over 7,000 hours and roughly 3M instruction-conditioned clips without preference or progress annotations. To make temporal-value learning reliable at scale, we combine random temporal sampling, temporal-order shuffling, and value-isolation attention, suppressing shortcuts that would leave predictions insensitive to failures and regressions. Trained without preference labels, RynnValue attains an average Kendall's tau_a of 0.675 on RBM-EVAL-OOD, surpassing the fully preference-supervised state of the art (0.655) and more than doubling a progress-only counterpart (0.292), while generalizing zero-shot to unseen tasks, embodiments, and viewpoints. Converted into dense rewards via potential-based shaping, it raises real-world policy success from 52.5% to 72.5% online and from 63.8% to 82.5% offline. These results establish temporal distance as a scalable supervision target and practical reward interface for generalist robot policies.
Original Article
View Cached Full Text

Cached at: 08/11/26, 06:19 AM

Paper page - RynnValue: Scaling Robotic Value Foundation Models with Temporal Distance

Source: https://huggingface.co/papers/2608.09853 Authors:

,

,

,

,

,

,

,

,

,

,

,

,

,

Abstract

General-purposerewardmodelsareincreasinglythebottleneckforscalingrobotlearning,yettherecipeforlearningvalue-relatedcapabilitiesfromlarge-scaleheterogeneouscorporaremainsunderexplored.Existingapproachestiesupervisiontotask-internalanchorssuchaspreferencesornormalizedprogress,noneofwhichtransfercleanlyacrossembodimentsanddatasources.WeintroduceRynnValue,anopen-sourcevaluefoundationmodelforroboticmanipulationthatreplacestheseanchorswithtemporaldistance,thedirectedcost-to-gofromanobservationtothelanguage-specifiedgoal.Becausetemporal-distancelabelscanbederiveddirectlyfromtimestamps,RynnValuescalestoover7,000hoursandroughly3Minstruction-conditionedclipswithoutpreferenceorprogressannotations.Tomaketemporal-valuelearningreliableatscale,wecombinerandomtemporalsampling,temporal-ordershuffling,andvalue-isolationattention,suppressingshortcutsthatwouldleavepredictionsinsensitivetofailuresandregressions.Trainedwithoutpreferencelabels,RynnValueattainsanaverageKendall’stau_aof0.675onRBM-EVAL-OOD,surpassingthefullypreference-supervisedstateoftheart(0.655)andmorethandoublingaprogress-onlycounterpart(0.292),whilegeneralizingzero-shottounseentasks,embodiments,andviewpoints.Convertedintodenserewardsviapotential-basedshaping,itraisesreal-worldpolicysuccessfrom52.5%to72.5%onlineandfrom63.8%to82.5%offline.Theseresultsestablishtemporaldistanceasascalablesupervisiontargetandpracticalrewardinterfaceforgeneralistrobotpolicies.

View arXiv pageView PDFProject pageGitHub7Add to collection

Get this paper in your agent:

hf papers read 2608\.09853

Don’t have the latest CLI?curl \-LsSf https://hf\.co/cli/install\.sh \| bash

Models citing this paper2

#### Alibaba-DAMO-Academy/RynnValue-4B Robotics• 5B• Updatedabout 1 hour ago • 11 • 3 #### Alibaba-DAMO-Academy/RynnValue-8B Robotics• 10B• Updatedabout 1 hour ago • 13 • 1

Datasets citing this paper0

No dataset linking this paper

Cite arxiv.org/abs/2608.09853 in a dataset README.md to link it from this page.

Spaces citing this paper0

No Space linking this paper

Cite arxiv.org/abs/2608.09853 in a Space README.md to link it from this page.

Collections including this paper1

Similar Articles

RynnWorld-4D: 4D Embodied World Models for Robotic Manipulation

Hugging Face Daily Papers

RynnWorld-4D is a generative world model that co-produces future RGB, depth, and optical flow from a single RGB-D image and language instruction using a unified diffusion process, enabling efficient robotic manipulation through inverse dynamics policy learning. It achieves state-of-the-art on real-world bimanual manipulation tasks.

World Value Models for Robotic Manipulation

Hugging Face Daily Papers

The paper presents World Value Model (WVM), a generalist robotic value model that combines world models with value estimation to accurately assess task progression and improve robotic policy learning from mixed-quality data, achieving state-of-the-art results on standard benchmarks and a new suboptimal data benchmark.

RoboTTT: Context Scaling for Robot Policies

Hugging Face Daily Papers

RoboTTT scales visuomotor context to 8K timesteps for robot policies, enabling one-shot imitation from human video demonstrations, on-the-fly policy improvement, and robustness to perturbations. It achieves an 87% improvement over baselines and completes a five-minute, ten-stage assembly task that no baseline could.