{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:MQVBUGG7KPL6ZTCUNMMAET77AV","short_pith_number":"pith:MQVBUGG7","schema_version":"1.0","canonical_sha256":"642a1a18df53d7eccc546b18024fff0557df1f30115d6b58e83eca24fcbe55c4","source":{"kind":"arxiv","id":"2401.04259","version":1},"attestation_state":"computed","paper":{"title":"MARG: Multi-Agent Review Generation for Scientific Papers","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Doug Downey, Larry Birnbaum, Mike D'Arcy, Tom Hope","submitted_at":"2024-01-08T22:24:17Z","abstract_excerpt":"We study the ability of LLMs to generate feedback for scientific papers and develop MARG, a feedback generation approach using multiple LLM instances that engage in internal discussion. By distributing paper text across agents, MARG can consume the full text of papers beyond the input length limitations of the base LLM, and by specializing agents and incorporating sub-tasks tailored to different comment types (experiments, clarity, impact) it improves the helpfulness and specificity of feedback. In a user study, baseline methods using GPT-4 were rated as producing generic or very generic comme"},"verification_status":{"content_addressed":true,"pith_receipt":true,"author_attested":false,"weak_author_claims":0,"strong_author_claims":0,"externally_anchored":false,"storage_verified":false,"citation_signatures":0,"replication_records":0,"graph_snapshot":true,"references_resolved":false,"formal_links_present":false},"canonical_record":{"source":{"id":"2401.04259","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-01-08T22:24:17Z","cross_cats_sorted":[],"title_canon_sha256":"53397e217a6eb2889912fc626103aa6c4a7df7d961208a29ae11a953bde9d2ad","abstract_canon_sha256":"3efca4e4ccf720b513090d454932697bd0fbf37c4c5c6f92521074fc55eba518"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T07:31:41.327936Z","signature_b64":"FUp8qYJ8++34VpJ4ONqLfJM67YBcY2IBx+BBUV0t5yArrzRWVrEptet0oJUOuDWLEdM7nAbWJXeoRJCVXELfAg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"642a1a18df53d7eccc546b18024fff0557df1f30115d6b58e83eca24fcbe55c4","last_reissued_at":"2026-07-05T07:31:41.327463Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T07:31:41.327463Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"MARG: Multi-Agent Review Generation for Scientific Papers","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Doug Downey, Larry Birnbaum, Mike D'Arcy, Tom Hope","submitted_at":"2024-01-08T22:24:17Z","abstract_excerpt":"We study the ability of LLMs to generate feedback for scientific papers and develop MARG, a feedback generation approach using multiple LLM instances that engage in internal discussion. By distributing paper text across agents, MARG can consume the full text of papers beyond the input length limitations of the base LLM, and by specializing agents and incorporating sub-tasks tailored to different comment types (experiments, clarity, impact) it improves the helpfulness and specificity of feedback. In a user study, baseline methods using GPT-4 were rated as producing generic or very generic comme"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2401.04259","kind":"arxiv","version":1},"verdict":{"id":null,"model_set":{},"created_at":null,"strongest_claim":"","one_line_summary":"","pipeline_version":null,"weakest_assumption":"","pith_extraction_headline":""},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2401.04259/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"},"aliases":[{"alias_kind":"arxiv","alias_value":"2401.04259","created_at":"2026-07-05T07:31:41.327522+00:00"},{"alias_kind":"arxiv_version","alias_value":"2401.04259v1","created_at":"2026-07-05T07:31:41.327522+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2401.04259","created_at":"2026-07-05T07:31:41.327522+00:00"},{"alias_kind":"pith_short_12","alias_value":"MQVBUGG7KPL6","created_at":"2026-07-05T07:31:41.327522+00:00"},{"alias_kind":"pith_short_16","alias_value":"MQVBUGG7KPL6ZTCU","created_at":"2026-07-05T07:31:41.327522+00:00"},{"alias_kind":"pith_short_8","alias_value":"MQVBUGG7","created_at":"2026-07-05T07:31:41.327522+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":16,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.25057","citing_title":"LLM-Based Scientific Peer Review: Methods, Benchmarks, and Reliability Challenges","ref_index":11,"is_internal_anchor":false},{"citing_arxiv_id":"2606.08857","citing_title":"PaperMentor: A Human-Centered Multi-Agent Writing Tutor for AI Research Papers on Overleaf","ref_index":20,"is_internal_anchor":false},{"citing_arxiv_id":"2606.05391","citing_title":"Human oversight of agentic systems in practice: Examining the oversight work, challenges, and heuristics of developers using software agents","ref_index":33,"is_internal_anchor":false},{"citing_arxiv_id":"2406.12708","citing_title":"AgentReview: Exploring Peer Review Dynamics with LLM Agents","ref_index":21,"is_internal_anchor":false},{"citing_arxiv_id":"2409.14634","citing_title":"Human-LLM Compound System for Scientific Ideation through Facet Recombination and Novelty Evaluation","ref_index":17,"is_internal_anchor":false},{"citing_arxiv_id":"2505.11336","citing_title":"XtraGPT: Context-Aware and Controllable Academic Paper Revision via Human-AI Collaboration","ref_index":51,"is_internal_anchor":false},{"citing_arxiv_id":"2605.18661","citing_title":"AI for Auto-Research: Roadmap & User Guide","ref_index":35,"is_internal_anchor":false},{"citing_arxiv_id":"2507.11810","citing_title":"Evolving Roles of LLMs in Scientific Innovation: Assistant, Collaborator, Scientist, and Evaluator","ref_index":24,"is_internal_anchor":false},{"citing_arxiv_id":"2411.04468","citing_title":"Magentic-One: A Generalist Multi-Agent System for Solving Complex Tasks","ref_index":7,"is_internal_anchor":false},{"citing_arxiv_id":"2601.03013","citing_title":"LLMs, You Can Evaluate It! Design of Multi-perspective Report Evaluation for Security Operation Centers","ref_index":23,"is_internal_anchor":false},{"citing_arxiv_id":"2604.04074","citing_title":"FactReview: Evidence-Grounded Reviews with Literature Positioning and Execution-Based Claim Verification","ref_index":2,"is_internal_anchor":false},{"citing_arxiv_id":"2501.06322","citing_title":"Multi-Agent Collaboration Mechanisms: A Survey of LLMs","ref_index":32,"is_internal_anchor":false},{"citing_arxiv_id":"2604.19578","citing_title":"Impact of large language models on peer review opinions from a fine-grained perspective: Evidence from top conference proceedings in AI","ref_index":42,"is_internal_anchor":false},{"citing_arxiv_id":"2408.06292","citing_title":"The AI Scientist: Towards Fully Automated Open-Ended Scientific Discovery","ref_index":19,"is_internal_anchor":false},{"citing_arxiv_id":"2604.06170","citing_title":"Paper Circle: An Open-source Multi-agent Research Discovery and Analysis Framework","ref_index":1,"is_internal_anchor":false},{"citing_arxiv_id":"2604.23593","citing_title":"When AI reviews science: Can we trust the referee?","ref_index":33,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/MQVBUGG7KPL6ZTCUNMMAET77AV","json":"https://pith.science/pith/MQVBUGG7KPL6ZTCUNMMAET77AV.json","graph_json":"https://pith.science/api/pith-number/MQVBUGG7KPL6ZTCUNMMAET77AV/graph.json","events_json":"https://pith.science/api/pith-number/MQVBUGG7KPL6ZTCUNMMAET77AV/events.json","paper":"https://pith.science/paper/MQVBUGG7"},"agent_actions":{"view_html":"https://pith.science/pith/MQVBUGG7KPL6ZTCUNMMAET77AV","download_json":"https://pith.science/pith/MQVBUGG7KPL6ZTCUNMMAET77AV.json","view_paper":"https://pith.science/paper/MQVBUGG7","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2401.04259&json=true","fetch_graph":"https://pith.science/api/pith-number/MQVBUGG7KPL6ZTCUNMMAET77AV/graph.json","fetch_events":"https://pith.science/api/pith-number/MQVBUGG7KPL6ZTCUNMMAET77AV/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/MQVBUGG7KPL6ZTCUNMMAET77AV/action/timestamp_anchor","attest_storage":"https://pith.science/pith/MQVBUGG7KPL6ZTCUNMMAET77AV/action/storage_attestation","attest_author":"https://pith.science/pith/MQVBUGG7KPL6ZTCUNMMAET77AV/action/author_attestation","sign_citation":"https://pith.science/pith/MQVBUGG7KPL6ZTCUNMMAET77AV/action/citation_signature","submit_replication":"https://pith.science/pith/MQVBUGG7KPL6ZTCUNMMAET77AV/action/replication_record"}},"created_at":"2026-07-05T07:31:41.327522+00:00","updated_at":"2026-07-05T07:31:41.327522+00:00"}