{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2026:RITIFKOXBVM3H2XGQT4N23QQAZ","short_pith_number":"pith:RITIFKOX","canonical_record":{"source":{"id":"2602.06239","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-02-05T22:31:07Z","cross_cats_sorted":[],"title_canon_sha256":"db7ddacc3944814b6e89500111aa47ddc7ffbdfc4cd5b094ac1b2867786e411b","abstract_canon_sha256":"138b14a500ed947f80d010919a83f4bdc3475151f49ad3e001a915a05d5c0d43"},"schema_version":"1.0"},"canonical_sha256":"8a2682a9d70d59b3eae684f8dd6e1006674a20be9c82ad8450a542a6432ce51d","source":{"kind":"arxiv","id":"2602.06239","version":2},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2602.06239","created_at":"2026-05-18T02:44:31Z"},{"alias_kind":"arxiv_version","alias_value":"2602.06239v2","created_at":"2026-05-18T02:44:31Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2602.06239","created_at":"2026-05-18T02:44:31Z"},{"alias_kind":"pith_short_12","alias_value":"RITIFKOXBVM3","created_at":"2026-05-18T12:33:37Z"},{"alias_kind":"pith_short_16","alias_value":"RITIFKOXBVM3H2XG","created_at":"2026-05-18T12:33:37Z"},{"alias_kind":"pith_short_8","alias_value":"RITIFKOX","created_at":"2026-05-18T12:33:37Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2026:RITIFKOXBVM3H2XGQT4N23QQAZ","target":"record","payload":{"canonical_record":{"source":{"id":"2602.06239","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-02-05T22:31:07Z","cross_cats_sorted":[],"title_canon_sha256":"db7ddacc3944814b6e89500111aa47ddc7ffbdfc4cd5b094ac1b2867786e411b","abstract_canon_sha256":"138b14a500ed947f80d010919a83f4bdc3475151f49ad3e001a915a05d5c0d43"},"schema_version":"1.0"},"canonical_sha256":"8a2682a9d70d59b3eae684f8dd6e1006674a20be9c82ad8450a542a6432ce51d","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-18T02:44:31.368909Z","signature_b64":"EFk47imAmxA499dUTgm7++GstpNtyV3oZtBwJDStrXh6w5OLIWWSy6hP4oRGmftKIcv2FbVE6iYCnmUNwfwYCg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"8a2682a9d70d59b3eae684f8dd6e1006674a20be9c82ad8450a542a6432ce51d","last_reissued_at":"2026-05-18T02:44:31.368347Z","signature_status":"signed_v1","first_computed_at":"2026-05-18T02:44:31.368347Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2602.06239","source_version":2,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-05-18T02:44:31Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"t00uHBffh6c200iTHWCqBrCMMTOC1EKJpOktyiHFuaAZjKD5wrYOgcCkq3pMJEpO2K9g/7KBr8ytU80VHWEWDw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-05-22T11:02:43.970871Z"},"content_sha256":"382464395d9fd28f60817edc318bf6f65c08470a1d814b80aca816db132a39c6","schema_version":"1.0","event_id":"sha256:382464395d9fd28f60817edc318bf6f65c08470a1d814b80aca816db132a39c6"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2026:RITIFKOXBVM3H2XGQT4N23QQAZ","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Provably avoiding over-optimization in Direct Preference Optimization without knowing the data distribution","license":"http://creativecommons.org/licenses/by/4.0/","headline":"PEPO mitigates DPO over-optimization by achieving sample complexity bounds that depend only on single-policy concentrability.","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Adam Barla, Emanuele Nevali, Luca Viano, Volkan Cevher","submitted_at":"2026-02-05T22:31:07Z","abstract_excerpt":"We introduce PEPO (Pessimistic Ensemble based Preference Optimization), a single-step Direct Preference Optimization (DPO)-like algorithm to mitigate the well-known over-optimization issue in preference learning without requiring the knowledge of the data-generating distribution or learning an explicit reward model. PEPO achieves pessimism via an ensemble of preference-optimized policies trained on disjoint data subsets and then aggregates them through a worst case construction that favors the agreement across models. In the tabular setting, PEPO achieves sample complexity guarantees depending"},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"In the tabular setting, PEPO achieves sample complexity guarantees depending only on a single-policy concentrability coefficient, thus avoiding the all-policy concentrability which affects the guarantees of algorithms prone to over-optimization, such as DPO.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"That an ensemble of policies trained on disjoint subsets can be aggregated via worst-case construction to produce pessimism without access to the data-generating distribution or explicit reward model.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"PEPO uses pessimistic ensembling of DPO policies on data subsets to achieve single-policy concentrability sample bounds and avoid over-optimization in tabular settings.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"PEPO mitigates DPO over-optimization by achieving sample complexity bounds that depend only on single-policy concentrability.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"9e2f8d8968b7d60ffb3f409dec81c4d8529ae42b5722e3dbf8ae85e7fb8c9da1"},"source":{"id":"2602.06239","kind":"arxiv","version":2},"verdict":{"id":"8e62023a-b6ba-4e51-9bc1-9df721e97474","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-16T06:32:34.967797Z","strongest_claim":"In the tabular setting, PEPO achieves sample complexity guarantees depending only on a single-policy concentrability coefficient, thus avoiding the all-policy concentrability which affects the guarantees of algorithms prone to over-optimization, such as DPO.","one_line_summary":"PEPO uses pessimistic ensembling of DPO policies on data subsets to achieve single-policy concentrability sample bounds and avoid over-optimization in tabular settings.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"That an ensemble of policies trained on disjoint subsets can be aggregated via worst-case construction to produce pessimism without access to the data-generating distribution or explicit reward model.","pith_extraction_headline":"PEPO mitigates DPO over-optimization by achieving sample complexity bounds that depend only on single-policy concentrability."},"references":{"count":49,"sample":[{"doi":"","year":null,"title":"Design considerations in offline preference-based rl","work_id":"09c3c6cb-11d2-44d6-a1ad-296f9a473b37","ref_index":1,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":null,"title":"XRPO: Pushing the limits of GRPO with targeted exploration and exploitation","work_id":"c1e825b8-b415-4c32-a6a6-b434dbb9a276","ref_index":2,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":null,"title":"Value-incentivized preference optimization: A unified approach to online and offline rlhf.arXiv preprint arXiv:2405.19320,","work_id":"6d4c354e-e135-4ac0-8b34-169bdac4b908","ref_index":3,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":null,"title":"On extending direct preference optimization to accommodate ties.arXiv preprint arXiv:2409.17431,","work_id":"0068f7ac-2b68-441b-8d52-de280b26eb0e","ref_index":4,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":null,"title":"AvoidingO(eRmax)scaling in rlhf through preference-based exploration.arXiv preprint arXiv:2502.00666,","work_id":"3c83224d-7ff1-4379-960c-e55e2b2fe73f","ref_index":5,"cited_arxiv_id":"","is_internal_anchor":false}],"resolved_work":49,"snapshot_sha256":"07dc1bd4215a1b76fb1d112528984e69821ad42546df3e7c957605760659cb0b","internal_anchors":5},"formal_canon":{"evidence_count":2,"snapshot_sha256":"dc323d102f41ea093701f3a0eef8044f6c409d8753281be99a33d9fd22e68810"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"},"verdict_id":"8e62023a-b6ba-4e51-9bc1-9df721e97474"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-05-18T02:44:31Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"Rhu5tQS8MhPbJJ7KslqML4BFlRz1xG0R+zRIdVEVA2myteYyqUK1bNsTeV8tDmRd7pHVqlAE1wgqT54COva6BA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-05-22T11:02:43.971843Z"},"content_sha256":"fe866cd6453fd1b5f136fdc5d8624008f10bb909e950766e850d513e08e51e72","schema_version":"1.0","event_id":"sha256:fe866cd6453fd1b5f136fdc5d8624008f10bb909e950766e850d513e08e51e72"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/RITIFKOXBVM3H2XGQT4N23QQAZ/bundle.json","state_url":"https://pith.science/pith/RITIFKOXBVM3H2XGQT4N23QQAZ/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/RITIFKOXBVM3H2XGQT4N23QQAZ/bundle.json","status":"primary"}],"public_keys":[{"key_id":"pith-v1-2026-05","algorithm":"ed25519","format":"raw","public_key_b64":"stVStoiQhXFxp4s2pdzPNoqVNBMojDU/fJ2db5S3CbM=","public_key_hex":"b2d552b68890857171a78b36a5dccf368a953413288c353f7c9d9d6f94b709b3","fingerprint_sha256_b32_first128bits":"RVFV5Z2OI2J3ZUO7ERDEBCYNKS","fingerprint_sha256_hex":"8d4b5ee74e4693bcd1df2446408b0d54","rotates_at":null,"url":"https://pith.science/pith-signing-key.json","notes":"Pith uses this Ed25519 key to sign canonical record SHA-256 digests. Verify with: ed25519_verify(public_key, message=canonical_sha256_bytes, signature=base64decode(signature_b64))."}],"merge_version":"pith-open-graph-merge-v1","built_at":"2026-05-22T11:02:43Z","links":{"resolver":"https://pith.science/pith/RITIFKOXBVM3H2XGQT4N23QQAZ","bundle":"https://pith.science/pith/RITIFKOXBVM3H2XGQT4N23QQAZ/bundle.json","state":"https://pith.science/pith/RITIFKOXBVM3H2XGQT4N23QQAZ/state.json","well_known_bundle":"https://pith.science/.well-known/pith/RITIFKOXBVM3H2XGQT4N23QQAZ/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:RITIFKOXBVM3H2XGQT4N23QQAZ","merge_version":"pith-open-graph-merge-v1","event_count":2,"valid_event_count":2,"invalid_event_count":0,"equivocation_count":0,"current":{"canonical_record":{"metadata":{"abstract_canon_sha256":"138b14a500ed947f80d010919a83f4bdc3475151f49ad3e001a915a05d5c0d43","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-02-05T22:31:07Z","title_canon_sha256":"db7ddacc3944814b6e89500111aa47ddc7ffbdfc4cd5b094ac1b2867786e411b"},"schema_version":"1.0","source":{"id":"2602.06239","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2602.06239","created_at":"2026-05-18T02:44:31Z"},{"alias_kind":"arxiv_version","alias_value":"2602.06239v2","created_at":"2026-05-18T02:44:31Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2602.06239","created_at":"2026-05-18T02:44:31Z"},{"alias_kind":"pith_short_12","alias_value":"RITIFKOXBVM3","created_at":"2026-05-18T12:33:37Z"},{"alias_kind":"pith_short_16","alias_value":"RITIFKOXBVM3H2XG","created_at":"2026-05-18T12:33:37Z"},{"alias_kind":"pith_short_8","alias_value":"RITIFKOX","created_at":"2026-05-18T12:33:37Z"}],"graph_snapshots":[{"event_id":"sha256:fe866cd6453fd1b5f136fdc5d8624008f10bb909e950766e850d513e08e51e72","target":"graph","created_at":"2026-05-18T02:44:31Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"graph_snapshot":{"author_claims":{"count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","strong_count":0},"builder_version":"pith-number-builder-2026-05-17-v1","claims":{"count":4,"items":[{"attestation":"unclaimed","claim_id":"C1","kind":"strongest_claim","source":"verdict.strongest_claim","status":"machine_extracted","text":"In the tabular setting, PEPO achieves sample complexity guarantees depending only on a single-policy concentrability coefficient, thus avoiding the all-policy concentrability which affects the guarantees of algorithms prone to over-optimization, such as DPO."},{"attestation":"unclaimed","claim_id":"C2","kind":"weakest_assumption","source":"verdict.weakest_assumption","status":"machine_extracted","text":"That an ensemble of policies trained on disjoint subsets can be aggregated via worst-case construction to produce pessimism without access to the data-generating distribution or explicit reward model."},{"attestation":"unclaimed","claim_id":"C3","kind":"one_line_summary","source":"verdict.one_line_summary","status":"machine_extracted","text":"PEPO uses pessimistic ensembling of DPO policies on data subsets to achieve single-policy concentrability sample bounds and avoid over-optimization in tabular settings."},{"attestation":"unclaimed","claim_id":"C4","kind":"headline","source":"verdict.pith_extraction.headline","status":"machine_extracted","text":"PEPO mitigates DPO over-optimization by achieving sample complexity bounds that depend only on single-policy concentrability."}],"snapshot_sha256":"9e2f8d8968b7d60ffb3f409dec81c4d8529ae42b5722e3dbf8ae85e7fb8c9da1"},"formal_canon":{"evidence_count":2,"snapshot_sha256":"dc323d102f41ea093701f3a0eef8044f6c409d8753281be99a33d9fd22e68810"},"paper":{"abstract_excerpt":"We introduce PEPO (Pessimistic Ensemble based Preference Optimization), a single-step Direct Preference Optimization (DPO)-like algorithm to mitigate the well-known over-optimization issue in preference learning without requiring the knowledge of the data-generating distribution or learning an explicit reward model. PEPO achieves pessimism via an ensemble of preference-optimized policies trained on disjoint data subsets and then aggregates them through a worst case construction that favors the agreement across models. In the tabular setting, PEPO achieves sample complexity guarantees depending","authors_text":"Adam Barla, Emanuele Nevali, Luca Viano, Volkan Cevher","cross_cats":[],"headline":"PEPO mitigates DPO over-optimization by achieving sample complexity bounds that depend only on single-policy concentrability.","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-02-05T22:31:07Z","title":"Provably avoiding over-optimization in Direct Preference Optimization without knowing the data distribution"},"references":{"count":49,"internal_anchors":5,"resolved_work":49,"sample":[{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":1,"title":"Design considerations in offline preference-based rl","work_id":"09c3c6cb-11d2-44d6-a1ad-296f9a473b37","year":null},{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":2,"title":"XRPO: Pushing the limits of GRPO with targeted exploration and exploitation","work_id":"c1e825b8-b415-4c32-a6a6-b434dbb9a276","year":null},{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":3,"title":"Value-incentivized preference optimization: A unified approach to online and offline rlhf.arXiv preprint arXiv:2405.19320,","work_id":"6d4c354e-e135-4ac0-8b34-169bdac4b908","year":null},{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":4,"title":"On extending direct preference optimization to accommodate ties.arXiv preprint arXiv:2409.17431,","work_id":"0068f7ac-2b68-441b-8d52-de280b26eb0e","year":null},{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":5,"title":"AvoidingO(eRmax)scaling in rlhf through preference-based exploration.arXiv preprint arXiv:2502.00666,","work_id":"3c83224d-7ff1-4379-960c-e55e2b2fe73f","year":null}],"snapshot_sha256":"07dc1bd4215a1b76fb1d112528984e69821ad42546df3e7c957605760659cb0b"},"source":{"id":"2602.06239","kind":"arxiv","version":2},"verdict":{"created_at":"2026-05-16T06:32:34.967797Z","id":"8e62023a-b6ba-4e51-9bc1-9df721e97474","model_set":{"reader":"grok-4.3"},"one_line_summary":"PEPO uses pessimistic ensembling of DPO policies on data subsets to achieve single-policy concentrability sample bounds and avoid over-optimization in tabular settings.","pipeline_version":"pith-pipeline@v0.9.0","pith_extraction_headline":"PEPO mitigates DPO over-optimization by achieving sample complexity bounds that depend only on single-policy concentrability.","strongest_claim":"In the tabular setting, PEPO achieves sample complexity guarantees depending only on a single-policy concentrability coefficient, thus avoiding the all-policy concentrability which affects the guarantees of algorithms prone to over-optimization, such as DPO.","weakest_assumption":"That an ensemble of policies trained on disjoint subsets can be aggregated via worst-case construction to produce pessimism without access to the data-generating distribution or explicit reward model."}},"verdict_id":"8e62023a-b6ba-4e51-9bc1-9df721e97474"}}],"author_attestations":[],"timestamp_anchors":[],"storage_attestations":[],"citation_signatures":[],"replication_records":[],"corrections":[],"mirror_hints":[],"record_created":{"event_id":"sha256:382464395d9fd28f60817edc318bf6f65c08470a1d814b80aca816db132a39c6","target":"record","created_at":"2026-05-18T02:44:31Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"attestation_state":"computed","canonical_record":{"metadata":{"abstract_canon_sha256":"138b14a500ed947f80d010919a83f4bdc3475151f49ad3e001a915a05d5c0d43","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-02-05T22:31:07Z","title_canon_sha256":"db7ddacc3944814b6e89500111aa47ddc7ffbdfc4cd5b094ac1b2867786e411b"},"schema_version":"1.0","source":{"id":"2602.06239","kind":"arxiv","version":2}},"canonical_sha256":"8a2682a9d70d59b3eae684f8dd6e1006674a20be9c82ad8450a542a6432ce51d","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"8a2682a9d70d59b3eae684f8dd6e1006674a20be9c82ad8450a542a6432ce51d","first_computed_at":"2026-05-18T02:44:31.368347Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-05-18T02:44:31.368347Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"EFk47imAmxA499dUTgm7++GstpNtyV3oZtBwJDStrXh6w5OLIWWSy6hP4oRGmftKIcv2FbVE6iYCnmUNwfwYCg==","signature_status":"signed_v1","signed_at":"2026-05-18T02:44:31.368909Z","signed_message":"canonical_sha256_bytes"},"source_id":"2602.06239","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:382464395d9fd28f60817edc318bf6f65c08470a1d814b80aca816db132a39c6","sha256:fe866cd6453fd1b5f136fdc5d8624008f10bb909e950766e850d513e08e51e72"],"state_sha256":"2674f67455b29f225169ac2e10294cc71000de452e633b83451297cf61c924d4"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"yNHa70CjvyYwUgFlr6eeYdBt8IUum9f9T+zkQ7QSR77esVC7Iib3NILpPz1skHUKnldYuNCupd7WULS7O9xPAQ==","signed_message":"bundle_sha256_bytes","signed_at":"2026-05-22T11:02:43.975846Z","bundle_sha256":"c16011086c232c55cb25ae1b0a0cc6eadd10965f68700cf2dc44fe35779637e8"}}