TRACE: A Unified Rollout Budget Allocation Framework for Efficient Agentic Reinforcement Learning

Hugging Face Daily Papers Papers

Summary

TRACE is a unified rollout budget allocation framework that enhances reward contrast in multi-turn agentic reinforcement learning by dynamically distributing resources across tree-structured rollouts based on prefix-level informativeness. It improves efficiency and accuracy on agentic benchmarks like Multi-Hop QA.

Reinforcement learning with verifiable rewards (RLVR) is a promising approach for enhancing reasoning and agentic behavior in large language models. However, rollout-intensive policy optimization is often limited by insufficient reward contrast, arising when overly simple or complex prompts generate low-variance feedback and when outcome-only rewards assign the same terminal assessment to every decision in a multi-turn rollout. Past efforts have focused on allocating available rollout resources to promising prompts, yet they only leverage sample informativeness at the prompt level and neglect variation in prefix-level informativeness across turns within the same rollout. This work targets multi-turn agentic RL by modeling each ReAct-style thought-action-observation turn as a semantically distinct node, allowing budget allocation to extend from prompt roots to turn-level prefixes with further continuations, which naturally forms tree-structured rollouts. We introduce Tree Rollout Allocation for Contrastive Exploration (TRACE), a unified rollout allocation framework that enhances reward contrast within a fixed sampling budget. Technically, TRACE allocates rollout budget to both prompt roots and intermediate prefixes that are most likely to yield mixed terminal rewards. A shared generalizable predictor estimates conditional success probability at these anchors from prefix histories to guide this allocation. The resulting adaptive tree structure enriches outcome-only feedback and amplifies the policy-update signal. Empirically, TRACE achieves competitive performance and efficiency gains on typical agentic benchmarks, e.g., improving Qwen3-14B Multi-Hop QA average accuracy by 2.8 points over competitive baselines at equal sampling cost.
Original Article
View Cached Full Text

Cached at: 06/11/26, 01:37 PM

Paper page - TRACE: A Unified Rollout Budget Allocation Framework for Efficient Agentic Reinforcement Learning

Source: https://huggingface.co/papers/2606.11119 Authors:

,

,

,

,

,

,

,

,

,

,

Abstract

TRACE is a rollout allocation framework that improves reward contrast in multi-turn agentic reinforcement learning by dynamically distributing resources across tree-structured rollouts based on prefix-level informativeness.

Reinforcement learningwithverifiable rewards(RLVR) is a promising approach for enhancing reasoning and agentic behavior in large language models. However, rollout-intensivepolicy optimizationis often limited by insufficientreward contrast, arising when overly simple or complex prompts generate low-variance feedback and when outcome-only rewards assign the same terminal assessment to every decision in a multi-turn rollout. Past efforts have focused on allocating available rollout resources to promising prompts, yet they only leverage sample informativeness at the prompt level and neglect variation in prefix-level informativeness across turns within the same rollout. This work targets multi-turn agentic RL by modeling eachReAct-stylethought-action-observation turn as a semantically distinct node, allowing budget allocation to extend from prompt roots to turn-level prefixes with further continuations, which naturally formstree-structured rollouts. We introduce TreeRollout Allocationfor Contrastive Exploration (TRACE), a unifiedrollout allocationframework that enhancesreward contrastwithin a fixed sampling budget. Technically, TRACE allocates rollout budget to both prompt roots and intermediate prefixes that are most likely to yield mixed terminal rewards. A shared generalizable predictor estimatesconditional success probabilityat these anchors from prefix histories to guide this allocation. The resulting adaptive tree structure enriches outcome-only feedback and amplifies thepolicy-update signal. Empirically, TRACE achieves competitive performance and efficiency gains on typical agentic benchmarks, e.g., improving Qwen3-14B Multi-Hop QA average accuracy by 2.8 points over competitive baselines at equal sampling cost.

View arXiv pageView PDFAdd to collection

Get this paper in your agent:

hf papers read 2606\.11119

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/2606.11119 in a model README.md to link it from this page.

Datasets citing this paper0

No dataset linking this paper

Cite arxiv.org/abs/2606.11119 in a dataset README.md to link it from this page.

Spaces citing this paper0

No Space linking this paper

Cite arxiv.org/abs/2606.11119 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

Cross-Epoch Adaptive Rollout Optimization for RL Post-Training

arXiv cs.LG

This paper presents CERO, a cross-epoch adaptive rollout optimization method for RL post-training of LLMs, which allocates a fixed rollout budget across prompts and epochs using Bayesian posterior variance to maximize sample efficiency, achieving theoretical regret bounds and outperforming GRPO on mathematical reasoning tasks.

Read the Trace, Steer the Path: Trajectory-Aware Reinforcement Learning for Diffusion Language Models

arXiv cs.CL

This paper introduces CAPR (Cached-Amortized Path Refinement), a reinforcement learning algorithm for diffusion large language models that extracts tree-like supervision signals from the denoising trace without the compute cost of full tree rollouts. CAPR achieves state-of-the-art performance on reasoning benchmarks like GSM8K, Math500, Sudoku, and Countdown at roughly 0.75x the cost of flat rollouts.