cd /news/artificial-intelligence/agentpatch-coarse-to-fine-weak-task-… · home topics artificial-intelligence article
[ARTICLE · art-89881] src=arxiv.org ↗ pub= topic=artificial-intelligence verified=true sentiment=· neutral

AgentPatch: Coarse-to-Fine Weak-Task Repair for Merging Agentic Multimodal Large Language Models

Researchers from an unnamed institution propose AgentPatch, a training-free coarse-to-fine repair framework that merges agentic multimodal large language models (MLLMs) into a single generalist checkpoint, addressing asymmetric capability preservation and behavior-critical forgetting. Experiments across six benchmarks show AgentPatch improves merged backbones, alleviates weak-task degradation, and balances recovery with preservation of complementary capabilities. Code is available at https://github.com/ziboshao/AgentPatch.

read1 min views1 publishedAug 10, 2026

arXiv:2608.06699v1 Announce Type: new Abstract: Agentic multimodal large language models (MLLMs) extend multimodal perception and reasoning with planning, tool use, and interaction in dynamic environments. Yet current models are specialized for particular tools or environments, complicating consolidation into a single generalist. We formulate Agentic MLLM Merging and identify two challenges: asymmetric capability preservation, whereby capabilities with different interaction complexity are retained unevenly, producing weak tasks after merging, and behavior-critical forgetting, whereby losing decisive actions can derail long-horizon execution. We propose AgentPatch, a training-free coarse-to-fine repair framework. It selects a stable merged backbone, restores diluted weak-task-specific signals through Weak-Task Unique Residual Recovery, and applies an Agent-Guided Behavior-Critical Patch that recovers decisive behaviors under explicit capability protection. AgentPatch produces a single static checkpoint without routing or ensembles. Experiments across six agentic and multimodal benchmarks show that AgentPatch improves diverse merged backbones, alleviates weak-task degradation, and better balances weak-task recovery with the preservation of complementary search and agentic visual processing capabilities. Code is available at https://github.com/ziboshao/AgentPatch.

── more in #artificial-intelligence 4 stories · sorted by recency
── more on @agentpatch 3 stories trending now
sponsored brought to you by zahid.host 4,200+ EU-deployed projects
reading about agents? ship yours in a single git push.

Run your AI side-project on zahid.host

EU-based hosting, git-push deploys, automatic HTTPS, no cold starts. Free tier with a custom domain — perfect for shipping the agent you just read about.

$git push zahid main
Live at https://your-agent.zahid.host
Get free account → Pricing
from €0/mo · no card required
LIVE [news/agentpatch-coarse-to…] indexed:0 read:1min 2026-08-10 ·