SVR-R1: Bootstrapping Multi-modal Reasoning with Self-verification in Reinforcement Learning

Hugging Face Daily Papers Papers

Summary

Introduces SVR-R1, a multi-turn reinforcement learning framework that uses the model's own verification as a learning signal for multi-modal reasoning, achieving significant accuracy improvements over standard GRPO baselines on vision-language reasoning benchmarks.

We introduce Self-Verified Reasoner (SVR-R1), a multi-turn RL framework that turns a model's own verification into a learning signal for multimodal reasoning. For each query, the model proposes an answer using the same weights, and issues a binary self-verdict (Yes/No). A 'No' triggers a second-chance rethink; a 'Yes,' or a turn cap, finalizes the output for computing the outcome-based reward. SVR-R1 is implemented with GRPO and an asynchronous multi-turn rollout framework and needs no external supervision or auxiliary critics. We evaluate SVR-R1 on vision-language reasoning benchmarks and show that it improves accuracy by a large margin over strong standard GRPO baselines. Training dynamics show decreasing reliance on verification-fewer verification turns, yet higher test accuracy-indicating that the gap between verification and generation narrows as the policy internalizes self-correction and chooses the most confident answer via our framework. SVR-R1 bridges the less explored intersection of inference-time self-refinement and RL training for VLMs, offering a simple yet effective recipe for bootstrapping multimodal reasoning. We will open-source SVR-R1 to facilitate future research in VLMs.
Original Article
View Cached Full Text

Cached at: 07/20/26, 09:26 PM

Paper page - SVR-R1: Bootstrapping Multi-modal Reasoning with Self-verification in Reinforcement Learning

Source: https://huggingface.co/papers/2607.10966 Authors:

,

,

,

,

,

,

,

,

,

,

,

,

Abstract

WeintroduceSelf-VerifiedReasoner(SVR-R1),amulti-turnRLframeworkthatturnsamodel’sownverificationintoalearningsignalformultimodalreasoning.Foreachquery,themodelproposesananswerusingthesameweights,andissuesabinaryself-verdict(Yes/No).A’No’triggersasecond-chancerethink;a’Yes,’oraturncap,finalizestheoutputforcomputingtheoutcome-basedreward.SVR-R1isimplementedwithGRPOandanasynchronousmulti-turnrolloutframeworkandneedsnoexternalsupervisionorauxiliarycritics.WeevaluateSVR-R1onvision-languagereasoningbenchmarksandshowthatitimprovesaccuracybyalargemarginoverstrongstandardGRPObaselines.Trainingdynamicsshowdecreasingrelianceonverification-fewerverificationturns,yethighertestaccuracy-indicatingthatthegapbetweenverificationandgenerationnarrowsasthepolicyinternalizesself-correctionandchoosesthemostconfidentanswerviaourframework.SVR-R1bridgesthelessexploredintersectionofinference-timeself-refinementandRLtrainingforVLMs,offeringasimpleyeteffectiverecipeforbootstrappingmultimodalreasoning.Wewillopen-sourceSVR-R1tofacilitatefutureresearchinVLMs.

View arXiv pageView PDFAdd to collection

Get this paper in your agent:

hf papers read 2607\.10966

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

Models citing this paper0

No model linking this paper

Cite arxiv.org/abs/2607.10966 in a model README.md to link it from this page.

Datasets citing this paper0

No dataset linking this paper

Cite arxiv.org/abs/2607.10966 in a dataset README.md to link it from this page.

Spaces citing this paper0

No Space linking this paper

Cite arxiv.org/abs/2607.10966 in a Space README.md to link it from this page.

Collections including this paper0

No Collection including this paper

Add this paper to acollectionto link it from this page.

Similar Articles

Reinforcing Dual-Path Reasoning in Spatial Vision Language Models

Hugging Face Daily Papers

This paper introduces SR-REAL, a unified framework for spatial vision-language models that combines linguistic deduction and 3D geometric reasoning via reinforcement learning, enabling robust multi-step spatial reasoning across diverse tasks.

Tandem Reinforcement Learning with Verifiable Rewards

arXiv cs.AI

Proposes Tandem Reinforcement Learning (TRL), extending the tandem training paradigm to RLVR to improve reasoning compatibility and legibility for weaker models and humans, showing that TRL matches solo performance while enhancing handoff robustness and reducing distributional drift.

Video Models Can Reason with Verifiable Rewards

Hugging Face Daily Papers

VideoRLVR optimizes video diffusion models for verifiable reasoning tasks using reinforcement learning with rule-based rewards, achieving better performance than supervised methods in constraint-satisfying video generation.