{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:BYWIXDXPYX3P5GBEGFIZD6VTN7","short_pith_number":"pith:BYWIXDXP","schema_version":"1.0","canonical_sha256":"0e2c8b8eefc5f6fe9824315191fab36fd459d40d1de2dc5f228c050d4bf961f1","source":{"kind":"arxiv","id":"2505.06548","version":1},"attestation_state":"computed","paper":{"title":"REFINE-AF: A Task-Agnostic Framework to Align Language Models via Self-Generated Instructions using Reinforcement Learning from Automated Feedback","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Abhilash Nandy, Aniruddha Roy, Pawan Goyal, Pretam Ray, Somak Aditya","submitted_at":"2025-05-10T07:23:19Z","abstract_excerpt":"Instruction-based Large Language Models (LLMs) have proven effective in numerous few-shot or zero-shot Natural Language Processing (NLP) tasks. However, creating human-annotated instruction data is time-consuming, expensive, and often limited in quantity and task diversity. Previous research endeavors have attempted to address this challenge by proposing frameworks capable of generating instructions in a semi-automated and task-agnostic manner directly from the model itself. Many of these efforts have relied on large API-only parameter-based models such as GPT-3.5 (175B), which are expensive, "},"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":"2505.06548","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-05-10T07:23:19Z","cross_cats_sorted":[],"title_canon_sha256":"f613ed1dc01160f2338bd34f1178e53b6e52008ce44baf912518012b6d014dd8","abstract_canon_sha256":"2bca8c6699f54aa83de8d923216275de50fac0e6222756d269466dc407dc69fb"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:01:12.015165Z","signature_b64":"e3R0C4CeqJSbWhpDvpnst7OdRz5JAKutFxF9BGtAGE/9kKZWwOtJKDeLGh4EbQf1ZGnvykiGu4W59OsZeCAjBg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"0e2c8b8eefc5f6fe9824315191fab36fd459d40d1de2dc5f228c050d4bf961f1","last_reissued_at":"2026-07-05T11:01:12.014685Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:01:12.014685Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"REFINE-AF: A Task-Agnostic Framework to Align Language Models via Self-Generated Instructions using Reinforcement Learning from Automated Feedback","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Abhilash Nandy, Aniruddha Roy, Pawan Goyal, Pretam Ray, Somak Aditya","submitted_at":"2025-05-10T07:23:19Z","abstract_excerpt":"Instruction-based Large Language Models (LLMs) have proven effective in numerous few-shot or zero-shot Natural Language Processing (NLP) tasks. However, creating human-annotated instruction data is time-consuming, expensive, and often limited in quantity and task diversity. Previous research endeavors have attempted to address this challenge by proposing frameworks capable of generating instructions in a semi-automated and task-agnostic manner directly from the model itself. Many of these efforts have relied on large API-only parameter-based models such as GPT-3.5 (175B), which are expensive, "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2505.06548","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/2505.06548/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":"2505.06548","created_at":"2026-07-05T11:01:12.014745+00:00"},{"alias_kind":"arxiv_version","alias_value":"2505.06548v1","created_at":"2026-07-05T11:01:12.014745+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2505.06548","created_at":"2026-07-05T11:01:12.014745+00:00"},{"alias_kind":"pith_short_12","alias_value":"BYWIXDXPYX3P","created_at":"2026-07-05T11:01:12.014745+00:00"},{"alias_kind":"pith_short_16","alias_value":"BYWIXDXPYX3P5GBE","created_at":"2026-07-05T11:01:12.014745+00:00"},{"alias_kind":"pith_short_8","alias_value":"BYWIXDXP","created_at":"2026-07-05T11:01:12.014745+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2604.16543","citing_title":"Conjunctive Prompt Attacks in Multi-Agent LLM Systems","ref_index":31,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/BYWIXDXPYX3P5GBEGFIZD6VTN7","json":"https://pith.science/pith/BYWIXDXPYX3P5GBEGFIZD6VTN7.json","graph_json":"https://pith.science/api/pith-number/BYWIXDXPYX3P5GBEGFIZD6VTN7/graph.json","events_json":"https://pith.science/api/pith-number/BYWIXDXPYX3P5GBEGFIZD6VTN7/events.json","paper":"https://pith.science/paper/BYWIXDXP"},"agent_actions":{"view_html":"https://pith.science/pith/BYWIXDXPYX3P5GBEGFIZD6VTN7","download_json":"https://pith.science/pith/BYWIXDXPYX3P5GBEGFIZD6VTN7.json","view_paper":"https://pith.science/paper/BYWIXDXP","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2505.06548&json=true","fetch_graph":"https://pith.science/api/pith-number/BYWIXDXPYX3P5GBEGFIZD6VTN7/graph.json","fetch_events":"https://pith.science/api/pith-number/BYWIXDXPYX3P5GBEGFIZD6VTN7/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/BYWIXDXPYX3P5GBEGFIZD6VTN7/action/timestamp_anchor","attest_storage":"https://pith.science/pith/BYWIXDXPYX3P5GBEGFIZD6VTN7/action/storage_attestation","attest_author":"https://pith.science/pith/BYWIXDXPYX3P5GBEGFIZD6VTN7/action/author_attestation","sign_citation":"https://pith.science/pith/BYWIXDXPYX3P5GBEGFIZD6VTN7/action/citation_signature","submit_replication":"https://pith.science/pith/BYWIXDXPYX3P5GBEGFIZD6VTN7/action/replication_record"}},"created_at":"2026-07-05T11:01:12.014745+00:00","updated_at":"2026-07-05T11:01:12.014745+00:00"}