{"paper":{"title":"SciDER: Scientific Data-centric End-to-end Researcher","license":"http://creativecommons.org/licenses/by/4.0/","headline":"SciDER deploys specialized agents to turn raw scientific data directly into hypotheses, experimental designs, and executable code.","cross_cats":["cs.CL"],"primary_cat":"cs.AI","authors_text":"Ke Lin, Owais Aijaz, Preslav Nakov, Xuehang Guo, Yilin Lu, Yiyang Luo","submitted_at":"2026-03-02T03:53:20Z","abstract_excerpt":"While large language models accelerate scientific discovery, existing agents face severe limitations in adaptability, domain generalization, and multimodal scalability, often struggling to autonomously process raw, domain-specific experimental data. To overcome these barriers, we introduce SciDER, a multi-agent system designed to flexibly automate the entire research lifecycle. This framework employs a novel data-centric approach and integrates a dynamic multimodal skill system across four specialized sub-agents. Specifically, an ideation agent generates novel hypotheses via Evolutionary Idea "},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"Evaluation on three benchmarks shows SciDER excels in specialized data-driven scientific discovery and outperforms general-purpose agents and state-of-the-art models through its self-evolving memory and critic-led feedback loop.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"That specialized collaborative agents can reliably parse and analyze arbitrary raw scientific data to produce hypotheses and experimental designs that are meaningfully grounded in the specific characteristics of that data.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"SciDER introduces a data-centric end-to-end multi-agent system that automates the scientific research lifecycle from raw data analysis to code execution and outperforms general-purpose agents on three benchmarks via self-evolving memory and critic feedback.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"SciDER deploys specialized agents to turn raw scientific data directly into hypotheses, experimental designs, and executable code.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"e20b4a4658b43a9308c106680bc4393bf2fb963ea606a30fb6f06f95465320ee"},"source":{"id":"2603.01421","kind":"arxiv","version":3},"verdict":{"id":"dd37f5c8-a4ec-4465-8dec-300a75b02890","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-15T18:40:30.768018Z","strongest_claim":"Evaluation on three benchmarks shows SciDER excels in specialized data-driven scientific discovery and outperforms general-purpose agents and state-of-the-art models through its self-evolving memory and critic-led feedback loop.","one_line_summary":"SciDER introduces a data-centric end-to-end multi-agent system that automates the scientific research lifecycle from raw data analysis to code execution and outperforms general-purpose agents on three benchmarks via self-evolving memory and critic feedback.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"That specialized collaborative agents can reliably parse and analyze arbitrary raw scientific data to produce hypotheses and experimental designs that are meaningfully grounded in the specific characteristics of that data.","pith_extraction_headline":"SciDER deploys specialized agents to turn raw scientific data directly into hypotheses, experimental designs, and executable code."},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2603.01421/integrity.json","findings":[],"available":true,"detectors_run":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938"},"references":{"count":0,"sample":[],"resolved_work":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","internal_anchors":0},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"}