SEAM learns to generate utility-optimized structured experiences via rollouts to boost frozen LLM performance on mathematical reasoning benchmarks with low overhead.
Mirac Suzgun, Mert Yuksekgonul, Federico Bianchi, Dan Jurafsky, and James Zou
9 Pith papers cite this work. Polarity classification is still indexing.
9
Pith papers citing it
citation-role summary
background 3
method 1
citation-polarity summary
years
2026 9representative citing papers
HypEHR is a hyperbolic embedding model for EHR data that uses Lorentzian geometry and hierarchy-aware pretraining to answer clinical questions nearly as well as large language models but with much smaller size.
A 4B deep research agent trained on 10K open data outperforms prior agents under 9B parameters and narrows the gap to 30B-class systems on research benchmarks.
The paper introduces the KDR task, HKA multi-agent framework, and KDR-Bench to enable LLM agents to integrate structured knowledge into deep research reports, with experiments showing outperformance over prior agents.
Multi-agent deep research systems self-optimize prompts through self-play to match or outperform expert-crafted versions.
Retrievers trained on agent trajectories via the LRAT framework improve evidence recall, task success, and efficiency in agentic search benchmarks.
Empirical study finds diversity collapse in multi-agent LLM ideation arises from structural coupling in interactions, not model limitations.
Pi-Serini shows a tuned BM25 lexical retriever with adequate depth, used inside an LLM agentic loop, reaches 83.1% accuracy and 94.7% evidence recall on BrowseComp-Plus while beating released dense-retriever agents.
Argues for a denoising-first paradigm in LLM-oriented information retrieval, framing challenges via a four-stage progression and providing a taxonomy of signal-to-noise optimization techniques across the pipeline.
citing papers explorer
-
Beyond Experience Retrieval: Learning to Generate Utility-Optimized Structured Experience for Frozen LLMs
SEAM learns to generate utility-optimized structured experiences via rollouts to boost frozen LLM performance on mathematical reasoning benchmarks with low overhead.
-
HypEHR: Hyperbolic Modeling of Electronic Health Records for Efficient Question Answering
HypEHR is a hyperbolic embedding model for EHR data that uses Lorentzian geometry and hierarchy-aware pretraining to answer clinical questions nearly as well as large language models but with much smaller size.
-
DR-Venus: Towards Frontier Edge-Scale Deep Research Agents with Only 10K Open Data
A 4B deep research agent trained on 10K open data outperforms prior agents under 9B parameters and narrows the gap to 30B-class systems on research benchmarks.
-
Towards Knowledgeable Deep Research: Framework and Benchmark
The paper introduces the KDR task, HKA multi-agent framework, and KDR-Bench to enable LLM agents to integrate structured knowledge into deep research reports, with experiments showing outperformance over prior agents.
-
Self-Optimizing Multi-Agent Systems for Deep Research
Multi-agent deep research systems self-optimize prompts through self-play to match or outperform expert-crafted versions.
-
Learning to Retrieve from Agent Trajectories
Retrievers trained on agent trajectories via the LRAT framework improve evidence recall, task success, and efficiency in agentic search benchmarks.
-
Diversity Collapse in Multi-Agent LLM Systems: Structural Coupling and Collective Failure in Open-Ended Idea Generation
Empirical study finds diversity collapse in multi-agent LLM ideation arises from structural coupling in interactions, not model limitations.
-
Rethinking Agentic Search with Pi-Serini: Is Lexical Retrieval Sufficient?
Pi-Serini shows a tuned BM25 lexical retriever with adequate depth, used inside an LLM agentic loop, reaches 83.1% accuracy and 94.7% evidence recall on BrowseComp-Plus while beating released dense-retriever agents.
-
LLM-Oriented Information Retrieval: A Denoising-First Perspective
Argues for a denoising-first paradigm in LLM-oriented information retrieval, framing challenges via a four-stage progression and providing a taxonomy of signal-to-noise optimization techniques across the pipeline.