From Search to Decision: How Post-Training Shapes Information-Seeking Capabilities in Large Language Models
cs.IR
Submitted: 2026-11-10
Updated: 2026-11-10
Terminology
Sources
- DRAGged into Conflicts: Detecting and Addressing Conflicting Sources in Search-Augmented LLMs
- QREF: Improving DeepSearch Agents by Aligning Search with Reasoning via Query Refinement and Evidence Filtering
- ToolSearcher: Optimizing Tool Selection at Scale via Reinforcement Learning
- Calibrate-Then-Act: Cost-Aware Exploration in LLM Agents
- Resolving Conflicting Evidence in Automated Fact-Checking: A Study on Retrieval-Augmented LLMs
- Optimizing Agentic Reasoning with Retrieval via Synthetic Semantic Information Gain Reward
- Reinforced Internal-External Knowledge Synergistic Reasoning for Efficient Adaptive Search Agent
- DeepRetrieval: Hacking Real Search Engines and Retrievers with Large Language Models via Reinforcement Learning
- Search-R1: Training LLMs to Reason and Leverage Search Engines with Reinforcement Learning
- Reinforcement Learning Foundations for Deep Research Systems: A Survey
- Towards Agentic RAG with Deep Reasoning: A Survey of RAG-Reasoning Systems in LLMs
- A Comprehensive Survey on Reinforcement Learning-based Agentic Search: Foundations, Roles, Optimizations, Evaluations, and Applications
- MARAG-R1: Beyond Single Retriever via Reinforcement-Learned Multi-Tool Agentic Retrieval
- Agentic Search in the Wild: Intents and Trajectory Dynamics from 14M+ Real Search Requests
- R1-Searcher: Incentivizing the Search Capability in LLMs via Reinforcement Learning
- ZeroSearch: Incentivize the Search Capability of LLMs without Searching
- One Policy, Any Budget: Internalizing Budget-Aware Search via Reinforcement Learning
- TRACE: Turn-level Reward Assignment via Credit Estimation for Long-Horizon Agents
- LongMemEval: Benchmarking Chat Assistants on Long-Term Interactive Memory
- Beyond Outcome Rewards: Step-Level Self-Distilled Policy Optimization for Deep Search Agents
Related papers
- UniRank: Benchmarking Ranking Models for Unified Sequential Modeling and Feature Interaction
- Revisiting Text Ranking in Deep Research
- GateSID: One Gate to Balance Semantic and Collaborative Signals across the Item Lifecycle
- SAM-D2Q: Aligning Multimodal Doc2Query with Search Demand and Conversion for E-commerce
- GINGER: Spicing Up Graph-Based Vector Search
- Subtraction Gets You More: Gap-Aware Retrieval for Multimodal Multi-Hop QA