Optimist Deep Research
An open-source iterative literature research agent system centered on an explicit research state. It supports sub-question tree decomposition, multi-source retrieval, citation snowball expansion, a structured evidence base and an eight-dimensional review mechanism, aiming to simulate the working style of human researchers.
Timeline (last 90 days)
Optimist Deep Research is an open-source project that proposes an iterative evidence-acquisition approach centered on an explicit, mutable research state.
Optimist Deep Research decomposes the research question into a tree of sub-questions, each driven by several search intents.
Optimist Deep Research's search intents are translated into specific queries targeting arXiv, OpenAlex, Crossref, PubMed, Semantic Scholar, and the open web.
Optimist Deep Research's retrieval results pass through a conservative four-tier deduplication hierarchy and are merged into a normalized document registry that retains source information.
Optimist Deep Research expands the document registry via citation snowballing until a saturation criterion is met.
Optimist Deep Research introduces a critic role that reviews the research state along eight dimensions and returns research actions rather than binary judgments.
The critic can add, split, rewrite, downgrade, or close sub-questions, making the research plan both an input to and an output of the investigation.
Optimist Deep Research's final report is written section by section based on the evidence base, and every citation can be traced back to an actually retrieved document.
In the case study reported by the authors, Optimist Deep Research retrieved 237 documents in about three and a half minutes and actually read 28 of them.
The authors acknowledge that the case study exposed weaknesses in Optimist Deep Research, showing that state-centric agent design is still at an early stage.
1 more timeline events
All Facts (11)
Optimist Deep Research is an open-source project that proposes an iterative evidence-acquisition approach centered on an explicit, mutable research state.
50%UnverifiedOptimist Deep Research decomposes the research question into a tree of sub-questions, each driven by several search intents.
50%UnverifiedOptimist Deep Research's search intents are translated into specific queries targeting arXiv, OpenAlex, Crossref, PubMed, Semantic Scholar, and the open web.
50%UnverifiedOptimist Deep Research's retrieval results pass through a conservative four-tier deduplication hierarchy and are merged into a normalized document registry that retains source information.
50%UnverifiedOptimist Deep Research expands the document registry via citation snowballing until a saturation criterion is met.
50%UnverifiedOptimist Deep Research introduces a critic role that reviews the research state along eight dimensions and returns research actions rather than binary judgments.
50%UnverifiedThe critic can add, split, rewrite, downgrade, or close sub-questions, making the research plan both an input to and an output of the investigation.
50%UnverifiedOptimist Deep Research's final report is written section by section based on the evidence base, and every citation can be traced back to an actually retrieved document.
50%UnverifiedIn the case study reported by the authors, Optimist Deep Research retrieved 237 documents in about three and a half minutes and actually read 28 of them.
50%UnverifiedThe authors acknowledge that the case study exposed weaknesses in Optimist Deep Research, showing that state-centric agent design is still at an early stage.
50%UnverifiedAutomatic assessment of evidence strength remains an open problem, and the current system may rely on heuristic rules or model scoring, whose reliability needs to be verified in more cases.
50%