SlimSearcher: Training Efficiency-Aware Web Agents via Adaptive Reward Gating

Hugging Face Daily Papers Papers
efficiency web-agents reinforcement-learning supervised-fine-tuning reward-shaping

Summary

SlimSearcher is a framework that improves efficiency in deep research agents by combining Pareto-efficient trajectory filtering and adaptive reward shaping, reducing tool-call rounds by 17-58% while maintaining accuracy on benchmarks like GAIA, BrowseComp, and XBenchDeepSearch.

Deep research agents have demonstrated remarkable capabilities in complex information-seeking tasks, yet this power comes at a steep computational cost. Driven by accuracy-focused training paradigms, current models adopt brute-force strategies characterized by blind tool dependency and performative reasoning-generating long, redundant trajectories that are far from necessary for resolving these tasks, leading to wasteful tool calls and excessive token consumption. To overcome this efficiency trap, we propose SlimSearcher, a principled framework that pushes the Pareto frontier between accuracy and computational cost across both Supervised Fine-Tuning (SFT) and Reinforcement Learning (RL). In the SFT stage, SlimSearcher employs Pareto-efficient filtration to distill trajectories that are both successful and economical, guiding the model toward inherently efficiency-aware search behaviors. During RL, we introduce Adaptive Reward Gating, a dynamic reward-shaping mechanism that evaluates relative tool and token efficiency within a sampled cohort. By cascading these adaptive efficiency metrics with a strict correctness gate, our approach effectively avoids the brevity bias associated with absolute penalties and mitigates reward hacking. Extensive experiments on long-horizon benchmarks, including GAIA, BrowseComp, and XBenchDeepSearch, demonstrate that SlimSearcher reduces average tool-call rounds by 17%-58% while maintaining or improving accuracy.
Original Article
View Cached Full Text

Cached at: 06/09/26, 12:42 PM

Paper page - SlimSearcher: Training Efficiency-Aware Web Agents via Adaptive Reward Gating

Source: https://huggingface.co/papers/2606.07074 Published on Jun 5

·

Submitted byhttps://huggingface.co/prayerdan

danon Jun 9

Abstract

SlimSearcher is a framework that improves efficiency in deep research agents by combining Pareto-efficient trajectory filtering and adaptive reward shaping to reduce computational costs while maintaining accuracy.

Deep research agents have demonstrated remarkable capabilities in complex information-seeking tasks, yet this power comes at a steep computational cost. Driven byaccuracy-focused training paradigms, current models adoptbrute-force strategiescharacterized by blind tool dependency andperformative reasoning-generating long, redundant trajectories that are far from necessary for resolving these tasks, leading to wasteful tool calls and excessive token consumption. To overcome this efficiency trap, we propose SlimSearcher, a principled framework that pushes the Pareto frontier between accuracy and computational cost across bothSupervised Fine-Tuning(SFT) andReinforcement Learning(RL). In the SFT stage, SlimSearcher employsPareto-efficient filtrationto distill trajectories that are both successful and economical, guiding the model toward inherently efficiency-aware search behaviors. During RL, we introduceAdaptive Reward Gating, a dynamicreward-shaping mechanismthat evaluates relative tool and token efficiency within a sampled cohort. By cascading these adaptive efficiency metrics with a strict correctness gate, our approach effectively avoids the brevity bias associated with absolute penalties and mitigates reward hacking. Extensive experiments on long-horizon benchmarks, including GAIA, BrowseComp, and XBenchDeepSearch, demonstrate that SlimSearcher reduces averagetool-call roundsby 17%-58% while maintaining or improving accuracy.

View arXiv pageView PDFAdd to collection

Get this paper in your agent:

hf papers read 2606\.07074

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/2606.07074 in a model README.md to link it from this page.

Datasets citing this paper0

No dataset linking this paper

Cite arxiv.org/abs/2606.07074 in a dataset README.md to link it from this page.

Spaces citing this paper0

No Space linking this paper

Cite arxiv.org/abs/2606.07074 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

ARBOR: Online Process Rewards via a Reusable Rubric Buffer for Search Agents

arXiv cs.CL

ARBOR introduces a reusable rubric buffer to provide online process rewards for LLM-based search agents, improving training efficiency when outcome-only rewards are insufficient. It outperforms GRPO and DAPO on multi-hop QA benchmarks, converting up to 42% of zero-gradient training groups into informative ones.