Beacon: Knowing When and How to Perform Agentic Visual Reasoning

Hugging Face Daily Papers Papers
agentic-visual-reasoning multimodal-llm tool-use reinforcement-learning visual-reasoning research

Summary

This paper introduces Beacon, an agentic visual reasoning model that improves multimodal LLMs' ability to decide when to use tools and benefit from tool use, using Necessity-Aware Adaptive Reward and Hint-Guided Capability Expansion in reinforcement learning.

The fundamental goal of agentic visual reasoning is to improve the success rate of multimodal large language models (MLLMs) on complex tasks, rather than merely equipping them with a sophisticated yet inefficient reasoning paradigm. In this work, we rethink agentic visual reasoning through two key dimensions of tool use: Mode Adaptiveness (MA) and Tool Effect (TE). Mode Adaptiveness characterizes whether an MLLM can recognize when tools are truly necessary and invoke them accordingly, thereby avoiding unnecessary computational overhead while improving performance on challenging problems that require tool assistance. Tool Effect characterizes the actual impact of tool use: tools should extend the model's capabilities on problems unsolvable through text-only reasoning, while avoiding additional errors on problems that the model can already solve without tools. We conduct a comprehensive analysis to quantify these two properties and empirically reveal that existing agentic visual reasoning models exhibit limited Mode Adaptiveness, while the gains produced by tool use on hard examples are largely offset by the harm introduced on easy examples that the models can already solve. Motivated by these observations, we propose Beacon, a novel agentic visual reasoning model that achieves stronger overall performance, improved Mode Adaptiveness, and genuine tool-induced performance gains. At the core of Beacon are the Necessity-Aware Adaptive Reward and the Hint-Guided Capability Expansion mechanism in the reinforcement learning stage, which respectively encourage adaptive tool invocation based on task necessity and strengthen the model's tool-use capability on the most challenging problems. Extensive experiments across diverse benchmarks demonstrate the strong overall performance of Beacon and its substantial improvements in both Mode Adaptiveness and Tool Effect.
Original Article
View Cached Full Text

Cached at: 07/31/26, 05:53 AM

Paper page - Beacon: Knowing When and How to Perform Agentic Visual Reasoning

Source: https://huggingface.co/papers/2607.28595 Authors:

,

,

,

,

,

,

,

,

,

,

,

,

Abstract

Thefundamentalgoalofagenticvisualreasoningistoimprovethesuccessrateofmultimodallargelanguagemodels(MLLMs)oncomplextasks,ratherthanmerelyequippingthemwithasophisticatedyetinefficientreasoningparadigm.Inthiswork,werethinkagenticvisualreasoningthroughtwokeydimensionsoftooluse:ModeAdaptiveness(MA)andToolEffect(TE).ModeAdaptivenesscharacterizeswhetheranMLLMcanrecognizewhentoolsaretrulynecessaryandinvokethemaccordingly,therebyavoidingunnecessarycomputationaloverheadwhileimprovingperformanceonchallengingproblemsthatrequiretoolassistance.ToolEffectcharacterizestheactualimpactoftooluse:toolsshouldextendthemodel’scapabilitiesonproblemsunsolvablethroughtext-onlyreasoning,whileavoidingadditionalerrorsonproblemsthatthemodelcanalreadysolvewithouttools.WeconductacomprehensiveanalysistoquantifythesetwopropertiesandempiricallyrevealthatexistingagenticvisualreasoningmodelsexhibitlimitedModeAdaptiveness,whilethegainsproducedbytooluseonhardexamplesarelargelyoffsetbytheharmintroducedoneasyexamplesthatthemodelscanalreadysolve.Motivatedbytheseobservations,weproposeBeacon,anovelagenticvisualreasoningmodelthatachievesstrongeroverallperformance,improvedModeAdaptiveness,andgenuinetool-inducedperformancegains.AtthecoreofBeaconaretheNecessity-AwareAdaptiveRewardandtheHint-GuidedCapabilityExpansionmechanisminthereinforcementlearningstage,whichrespectivelyencourageadaptivetoolinvocationbasedontasknecessityandstrengthenthemodel’stool-usecapabilityonthemostchallengingproblems.ExtensiveexperimentsacrossdiversebenchmarksdemonstratethestrongoverallperformanceofBeaconanditssubstantialimprovementsinbothModeAdaptivenessandToolEffect.

View arXiv pageView PDFGitHub1Add to collection

Get this paper in your agent:

hf papers read 2607\.28595

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/2607.28595 in a model README.md to link it from this page.

Datasets citing this paper0

No dataset linking this paper

Cite arxiv.org/abs/2607.28595 in a dataset README.md to link it from this page.

Spaces citing this paper0

No Space linking this paper

Cite arxiv.org/abs/2607.28595 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

ATLAS: Agentic or Latent Visual Reasoning? One Word is Enough for Both

Hugging Face Daily Papers

ATLAS presents a visual reasoning framework that combines agentic operations and latent representations using functional tokens, enabling efficient training via next-token prediction and reinforcement learning while avoiding intermediate image generation.

Adaptive Latent Agentic Reasoning

arXiv cs.CL

This paper introduces Adaptive Latent Agentic Reasoning (ALAR), a dual-mode framework for LLM agents that uses compact latent reasoning for routine turns and selectively escalates to explicit chain-of-thought for harder decisions, achieving up to 84.6% token reduction while maintaining task accuracy.