Tag
This paper introduces AISE-Bench, a curated benchmark with 1,133 QA pairs for evaluating LLM agents on multi-step API planning and grounded summarization for academic knowledge graphs. The benchmark reveals that even the strongest model achieves only moderate performance, highlighting challenges in stepwise correctness and traceable reasoning.
Introduces SearchOS, a multi-agent framework for robust open-domain information-seeking that externalizes search progress into explicit states via a novel Search-Oriented Context Management (SOCM) system, achieving state-of-the-art results on WideSearch and GISA benchmarks.
The paper develops an agentic evaluation framework for clinical reasoning in hematologic oncology, finding that LLMs primarily fail due to systematic information-seeking deficits rather than insufficient knowledge, with error patterns resembling cognitive biases in novice clinicians.
This paper introduces a toy framework that models curiosity as an ecosystem in single and multi-agent settings, exploring how agents weigh immediate uncertainty reduction, costs, delayed returns, and the value of keeping questions open. It aims to inform future multi-agent AI systems for discovery.
This paper proposes ActionRating, a formulation that places clarification inside an agent's action space on a shared ordinal scale with navigation, enabling two information-seeking modes (mandatory and opportunistic). On hierarchical taxonomy classification benchmarks, experiments with 9 LLMs show that opportunistic clarification improves accuracy and information-seeking effectiveness.
Struct-Searcher introduces a belief revision theory-based structural agentic workflow for multimodal deep information seeking, achieving significant accuracy improvements over existing vision-language models and deep research agents.
WebShaper is a formalization-driven framework for synthesizing information-seeking datasets using set theory and Knowledge Projections, achieving state-of-the-art performance on GAIA and WebWalkerQA benchmarks among open-source agents.