The following papers were recommended by the Semantic Scholar API
\n- \n
- Search and Refine During Think: Autonomous Retrieval-Augmented Reasoning of LLMs (2025) \n
- R1-Searcher++: Incentivizing the Dynamic Knowledge Acquisition of LLMs via Reinforcement Learning (2025) \n
- Tool-Star: Empowering LLM-Brained Multi-Tool Reasoner via Reinforcement Learning (2025) \n
- StepSearch: Igniting LLMs Search Ability via Step-Wise Proximal Policy Optimization (2025) \n
- ZeroSearch: Incentivize the Search Capability of LLMs without Searching (2025) \n
- Learning to Route Queries Across Knowledge Bases for Step-wise Retrieval-Augmented Reasoning (2025) \n
- SimpleDeepSearcher: Deep Information Seeking via Web-Powered Reasoning Trajectory Synthesis (2025) \n
Please give a thumbs up to this comment if you found it helpful!
\nIf you want recommendations for any Paper on Hugging Face checkout this Space
\n You can directly ask Librarian Bot for paper recommendations by tagging it in a comment: \n\n@librarian-bot\n\t recommend
MaskSearch: A Universal Pre-Training Framework to Enhance Agentic Search Capability
Abstract
A novel pre-training framework, MaskSearch, enhances Large Language Models with universal retrieval and reasoning capabilities through a Retrieval Augmented Mask Prediction task, improving their performance in open-domain multi-hop question answering.
Retrieval-Augmented Language Models (RALMs) represent a classic paradigm where models enhance generative capabilities using external knowledge retrieved via a specialized module. Recent advancements in Agent techniques enable Large Language Models (LLMs) to autonomously utilize tools for retrieval, planning, and reasoning. While existing training-based methods show promise, their agentic abilities are limited by inherent characteristics of the task-specific data used during training. To further enhance the universal search capability of agents, we propose a novel pre-training framework, MaskSearch. In the pre-training stage, we introduce the Retrieval Augmented Mask Prediction (RAMP) task, where the model learns to leverage search tools to fill masked spans on a large number of pre-training data, thus acquiring universal retrieval and reasoning capabilities for LLMs. After that, the model is trained on downstream tasks to achieve further improvement. We apply both Supervised Fine-tuning (SFT) and Reinforcement Learning (RL) for training. For SFT, we combine agent-based and distillation-based methods to generate training data, starting with a multi-agent system consisting of a planner, rewriter, observer, and followed by a self-evolving teacher model. While for RL, we employ DAPO as the training framework and adopt a hybrid reward system consisting of answer rewards and format rewards. Additionally, we introduce a curriculum learning approach that allows the model to learn progressively from easier to more challenging instances based on the number of masked spans. We evaluate the effectiveness of our framework in the scenario of open-domain multi-hop question answering. Through extensive experiments, we demonstrate that MaskSearch significantly enhances the performance of LLM-based search agents on both in-domain and out-of-domain downstream tasks.
Community
Hi everyone! I'm super excited to dive into our advancement in the realm of agentic search capabilities of LLMs.
🌟 We propose MaskSearch, a novel pre-training framework to further enhance the universal search capability of agents.
📚 We introduce the Retrieval Augmented Mask Prediction (RAMP) task, where the model learns to leverage search tools to fill masked spans on a large number of pre-training data, thus acquiring universal retrieval and reasoning capabilities for LLMs.
🚀 We combine agent-based and distillation-based methods to generate training data, starting with a multi-agent system consisting of a planner, rewriter, observer, and followed by a self-evolving teacher model.
🎯 Extensive experiments demonstrate that MaskSearch significantly enhances the performance of LLM-based search agents on both in-domain and out-of-domain downstream tasks.
This is an automated message from the Librarian Bot. I found the following papers similar to this paper.
The following papers were recommended by the Semantic Scholar API
- Search and Refine During Think: Autonomous Retrieval-Augmented Reasoning of LLMs (2025)
- R1-Searcher++: Incentivizing the Dynamic Knowledge Acquisition of LLMs via Reinforcement Learning (2025)
- Tool-Star: Empowering LLM-Brained Multi-Tool Reasoner via Reinforcement Learning (2025)
- StepSearch: Igniting LLMs Search Ability via Step-Wise Proximal Policy Optimization (2025)
- ZeroSearch: Incentivize the Search Capability of LLMs without Searching (2025)
- Learning to Route Queries Across Knowledge Bases for Step-wise Retrieval-Augmented Reasoning (2025)
- SimpleDeepSearcher: Deep Information Seeking via Web-Powered Reasoning Trajectory Synthesis (2025)
Please give a thumbs up to this comment if you found it helpful!
If you want recommendations for any Paper on Hugging Face checkout this Space
You can directly ask Librarian Bot for paper recommendations by tagging it in a comment:
@librarian-bot
recommend
Models citing this paper 0
No model linking this paper
Datasets citing this paper 0
No dataset linking this paper
Spaces citing this paper 0
No Space linking this paper
Collections including this paper 0
No Collection including this paper