cd /news/artificial-intelligence/agent-explorative-policy-optimizatio… · home topics artificial-intelligence article
[ARTICLE · art-91660] src=research.nvidia.com ↗ pub= topic=artificial-intelligence verified=true sentiment=↑ positive

Agent Explorative Policy Optimization for Multimodal Agentic Reasoning

Researchers propose AXPO (Agent eXplorative Policy Optimization), a reinforcement learning method that addresses the Thinking-Acting Gap in multimodal agentic reasoning by resampling tool calls for all-wrong subgroups. Across nine multimodal benchmarks and three scales of Qwen3-VL-Thinking, SFT+AXPO outperforms SFT+GRPO by +1.8pp Pass@1 and +1.8pp Pass@4 on average at 8B, and the 8B model surpasses the 32B Base on Pass@4 with 4 times fewer parameters.

read1 min views1 publishedAug 11, 2026

Vision-language models with extended reasoning succeed on complex problems, but many real-world problems require external tools that internal reasoning alone often cannot resolve. Agentic reasoning therefore interleaves two behaviors with a structural asymmetry: thinking (the self-contained default) and tool use (a high-variance auxiliary acting). We refer to this asymmetry as the Thinking-Acting Gap. Under standard RL recipes like GRPO, the gap manifests as two diagnostic symptoms during training: tool use is attempted on only ~30% of rollouts, and when attempted, the tool-using rollouts within a group are all-wrong on ~40% of questions, suppressing the learning signal at the tool calls that needed it. We propose AXPO (Agent eXplorative Policy Optimization): for each all-wrong tool-using subgroup, AXPO fixes the thinking prefix and resamples the tool call and its continuation, paired with uncertainty-based prefix selection. Across nine multimodal benchmarks and three scales of Qwen3-VL-Thinking, SFT+AXPO outperforms SFT+GRPO at average (+1.8pp Pass@1 and +1.8pp Pass@4 at 8B on average) and 8B with SFT+AXPO surpasses the 32B Base on Pass@4 with 4 times fewer parameters.

── more in #artificial-intelligence 4 stories · sorted by recency
── more on @axpo 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/agent-explorative-po…] indexed:0 read:1min 2026-08-11 ·