REVIEW 15 cited by
WebDancer: Towards Autonomous Information Seeking Agency
Not yet reviewed by Pith; the record is open.
This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.
SPECIMEN: schema-true, not a live event
T0 review · schema-true
One-sentence machine reading of the paper's core claim.
pith:XXXXXXXX · record.json · timestamp
read the original abstract
Addressing intricate real-world problems necessitates in-depth information seeking and multi-step reasoning. Recent progress in agentic systems, exemplified by Deep Research, underscores the potential for autonomous multi-step research. In this work, we present a cohesive paradigm for building end-to-end agentic information seeking agents from a data-centric and training-stage perspective. Our approach consists of four key stages: (1) browsing data construction, (2) trajectories sampling, (3) supervised fine-tuning for effective cold start, and (4) reinforcement learning for enhanced generalisation. We instantiate this framework in a web agent based on the ReAct, WebDancer. Empirical evaluations on the challenging information seeking benchmarks, GAIA and WebWalkerQA, demonstrate the strong performance of WebDancer, achieving considerable results and highlighting the efficacy of our training paradigm. Further analysis of agent training provides valuable insights and actionable, systematic pathways for developing more capable agentic models. The codes and demo will be released in https://github.com/Alibaba-NLP/WebAgent.
Forward citations
Cited by 15 Pith papers
-
Video-DeepResearch: Towards the Next-Generation Multimodal Deepresearch Agent
Video-DeepResearch trains video agents with a two-stage SFT plus GRPO pipeline to do visual web searches before text searches, and introduces a 200-question benchmark requiring both video grounding and open-web knowledge.
-
From Simple QA to Deep Research: A Verifiable Benchmark Constructed through Iterative Task Evolution
An automatic pipeline evolves simple QA questions into 500 deep-research tasks with DAG-structured, fact-grounded rubrics that discriminate between models.
-
Fetch-then-Explore: Decoupling Selection from Extraction over a Persistent Workspace for Search Agents
Fetch-then-Explore, which stores fetched pages in a per-question workspace and extracts evidence on demand with grep/read, beats visit-and-read and browsing baselines on BrowseComp across three LLM backbones.
-
ToolAnchor: Anchoring Counterfactual Context to Boost Agentic Tool-use Capability
A three-stage teacher-hypothesize, student-verify, then-train pipeline lets a post-trained tool-using agent adopt new visual tools without retraining from scratch.
-
Entropy Pacing Policy Optimization for Multi-Task Agentic Reinforcement Learning
Replacing GRPO's fixed clipping range with a task-wise entropy-aware adaptive bound stabilizes multi-task agentic LLM training by synchronizing exploration-exploitation paces.
-
SearchEyes: Towards Frontier Multimodal Deep Search Intelligence via Search World Simulation
SearchEyes unifies multimodal search-agent training via Perception-Knowledge Chains on Wikidata5M and Hop-Anchored Policy Optimization, claiming a 6.2-point average gain over the strongest open-source baseline on six ...
-
MemSifter: Offloading LLM Memory Retrieval via Outcome-Driven Proxy Reasoning
MemSifter trains a 4B proxy with an outcome-driven, rank-sensitive RL reward to sift LLM memory, and on eight benchmarks it matches or beats embedding, graph, and long-context baselines.
-
Encouraging Good Processes Without the Need for Good Answers: Reinforcement Learning for LLM Agent Planning
Decoupling LLM-agent planning from summarization and rewarding tool-call completeness rather than final-answer correctness improves planning by 8-12% and end-to-end answers by 5-6% over end-to-end RL baselines.
-
Chain-of-Agents: End-to-End Agent Foundation Models via Multi-Agent Distillation and Agentic RL
A single LLM is trained with multi-agent distilled trajectories plus agentic RL, and the resulting Chain-of-Agents models set state-of-the-art Pass@1 scores among tool-integrated reasoning methods on GAIA, BrowseComp,...
-
MetaAgent: Toward Self-Evolving Agent via Tool Meta-Learning
The paper advertises a self-evolving tool-using agent that beats baselines, but the body contains a different math paper, leaving the performance claims unsupported.
-
WebShaper: Agentically Data Synthesizing via Information-Seeking Formalization
WebShaper formalizes web information-seeking tasks as set-theoretic queries, synthesizes training questions by layer-wise expansion, and uses the resulting data to train open-source agents that outperform prior open-s...
-
TAPO: Transition-Aware Policy Optimization for LLM Agents
Interleaving action-conditioned next-observation supervision with group RL on a shared LLM backbone consistently lifts long-horizon agent success over pure policy optimization.
-
LiteResearcher: A Scalable Agentic RL Training Framework for Deep Research Agent
LiteResearcher uses a lite virtual world to make agentic RL training scalable and stable, enabling a 4B model to achieve 71.3% on GAIA and 78.0% on Xbench, outperforming larger open-source and commercial systems.
-
SciMaster: Towards General-Purpose Scientific AI Agents, Part I. X-Master as Foundation: Can We Lead on Humanity's Last Exam?
X-Masters, an open-source scattered-and-stacked agent workflow built on DeepSeek-R1-0528, reports 32.1% on the HLE text-only subset, the first score above 30% and higher than OpenAI and Google Deep Research.
-
Deep Research Agents: A Systematic Examination And Roadmap
A survey that organizes LLM-powered deep research agents into static versus dynamic workflows and single versus multi agent architectures, and reviews their benchmarks and open challenges.
Discussion (0). Sign in to comment.