{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2026:QTQQZWERSLJH2JIZQXZG3TBIAU","short_pith_number":"pith:QTQQZWER","canonical_record":{"source":{"id":"2606.09043","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-06-08T05:24:15Z","cross_cats_sorted":["cs.CL"],"title_canon_sha256":"bd8e00f4557352d50810b826a5d2203c756c40d3cab08c77a8c93d8ab66c0c69","abstract_canon_sha256":"3a549ff70ec4e7237cf6f3cc9b1fc4bc3eca7d7c175da07fc01b0727f76d877c"},"schema_version":"1.0"},"canonical_sha256":"84e10cd89192d27d251985f26dcc280510705eb9bae68aee78ddd4279ae74c5f","source":{"kind":"arxiv","id":"2606.09043","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2606.09043","created_at":"2026-06-09T02:07:55Z"},{"alias_kind":"arxiv_version","alias_value":"2606.09043v1","created_at":"2026-06-09T02:07:55Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2606.09043","created_at":"2026-06-09T02:07:55Z"},{"alias_kind":"pith_short_12","alias_value":"QTQQZWERSLJH","created_at":"2026-06-09T02:07:55Z"},{"alias_kind":"pith_short_16","alias_value":"QTQQZWERSLJH2JIZ","created_at":"2026-06-09T02:07:55Z"},{"alias_kind":"pith_short_8","alias_value":"QTQQZWER","created_at":"2026-06-09T02:07:55Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2026:QTQQZWERSLJH2JIZQXZG3TBIAU","target":"record","payload":{"canonical_record":{"source":{"id":"2606.09043","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-06-08T05:24:15Z","cross_cats_sorted":["cs.CL"],"title_canon_sha256":"bd8e00f4557352d50810b826a5d2203c756c40d3cab08c77a8c93d8ab66c0c69","abstract_canon_sha256":"3a549ff70ec4e7237cf6f3cc9b1fc4bc3eca7d7c175da07fc01b0727f76d877c"},"schema_version":"1.0"},"canonical_sha256":"84e10cd89192d27d251985f26dcc280510705eb9bae68aee78ddd4279ae74c5f","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-09T02:07:55.295763Z","signature_b64":"xzMW1Fm4j01IxIk8WPS7+pxDQfB2W58bWXylZJVNN6wlrU5QsL4f23bPaD1RqjkrLvXnfVfP3n6GU4nm/Mt1Dw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"84e10cd89192d27d251985f26dcc280510705eb9bae68aee78ddd4279ae74c5f","last_reissued_at":"2026-06-09T02:07:55.294872Z","signature_status":"signed_v1","first_computed_at":"2026-06-09T02:07:55.294872Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2606.09043","source_version":1,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-06-09T02:07:55Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"/gUCmKTaw8vXNfEk3gZ2F/Jd4YLWnYN6FO2STLWShJjdKvg0nPcsA3dxPJdbKPxPqiAtoZ/g5Ec5QiRpmdocBg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-06-30T00:07:31.023008Z"},"content_sha256":"d6d7f20b582b1a09b8b24484457e75af7345c02cb8546d504ddbc35bf02be2dd","schema_version":"1.0","event_id":"sha256:d6d7f20b582b1a09b8b24484457e75af7345c02cb8546d504ddbc35bf02be2dd"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2026:QTQQZWERSLJH2JIZQXZG3TBIAU","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"DynaCF: Mitigating Shortcut Learning in Reward Models via Dynamic Counterfactual Sensitivity","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.CL"],"primary_cat":"cs.LG","authors_text":"Fei Sun, Fengyuan Liu, Mengnan Du, Yanguang Liu, Yongliang Miao, Zirui He","submitted_at":"2026-06-08T05:24:15Z","abstract_excerpt":"Reward models trained from pairwise preferences often exploit superficial shortcut cues rather than learning true response quality. We propose DynaCF, a dynamic reweighting framework for mitigating shortcut learning in reward model training. Unlike static shortcut heuristics, DynaCF measures shortcut sensitivity online during optimization by applying semantics-preserving counterfactual perturbations and tracking the resulting margin shifts and preference flips under the current model. Samples with higher shortcut sensitivity are dynamically downweighted in the Bradley-Terry objective, encourag"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2606.09043","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/2606.09043/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"},"verdict_id":null},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-06-09T02:07:55Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"5BVhI4Xbp1tTbP4UBTt7Ky7VeisMQn4+QF4OaSZB9KaDg8W2JpZG2R12FUpw5IWThsmqC5oM+83uA/Fda+0NCg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-06-30T00:07:31.023608Z"},"content_sha256":"7a36a5ba3a5762d1821ad0e47fae7e69c3d4a8c4543f49eeac7e5ae07446918a","schema_version":"1.0","event_id":"sha256:7a36a5ba3a5762d1821ad0e47fae7e69c3d4a8c4543f49eeac7e5ae07446918a"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/QTQQZWERSLJH2JIZQXZG3TBIAU/bundle.json","state_url":"https://pith.science/pith/QTQQZWERSLJH2JIZQXZG3TBIAU/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/QTQQZWERSLJH2JIZQXZG3TBIAU/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":"2026-06-30T00:07:31Z","links":{"resolver":"https://pith.science/pith/QTQQZWERSLJH2JIZQXZG3TBIAU","bundle":"https://pith.science/pith/QTQQZWERSLJH2JIZQXZG3TBIAU/bundle.json","state":"https://pith.science/pith/QTQQZWERSLJH2JIZQXZG3TBIAU/state.json","well_known_bundle":"https://pith.science/.well-known/pith/QTQQZWERSLJH2JIZQXZG3TBIAU/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:QTQQZWERSLJH2JIZQXZG3TBIAU","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":"3a549ff70ec4e7237cf6f3cc9b1fc4bc3eca7d7c175da07fc01b0727f76d877c","cross_cats_sorted":["cs.CL"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-06-08T05:24:15Z","title_canon_sha256":"bd8e00f4557352d50810b826a5d2203c756c40d3cab08c77a8c93d8ab66c0c69"},"schema_version":"1.0","source":{"id":"2606.09043","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2606.09043","created_at":"2026-06-09T02:07:55Z"},{"alias_kind":"arxiv_version","alias_value":"2606.09043v1","created_at":"2026-06-09T02:07:55Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2606.09043","created_at":"2026-06-09T02:07:55Z"},{"alias_kind":"pith_short_12","alias_value":"QTQQZWERSLJH","created_at":"2026-06-09T02:07:55Z"},{"alias_kind":"pith_short_16","alias_value":"QTQQZWERSLJH2JIZ","created_at":"2026-06-09T02:07:55Z"},{"alias_kind":"pith_short_8","alias_value":"QTQQZWER","created_at":"2026-06-09T02:07:55Z"}],"graph_snapshots":[{"event_id":"sha256:7a36a5ba3a5762d1821ad0e47fae7e69c3d4a8c4543f49eeac7e5ae07446918a","target":"graph","created_at":"2026-06-09T02:07:55Z","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":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"integrity":{"available":true,"clean":true,"detectors_run":[],"endpoint":"/pith/2606.09043/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Reward models trained from pairwise preferences often exploit superficial shortcut cues rather than learning true response quality. We propose DynaCF, a dynamic reweighting framework for mitigating shortcut learning in reward model training. Unlike static shortcut heuristics, DynaCF measures shortcut sensitivity online during optimization by applying semantics-preserving counterfactual perturbations and tracking the resulting margin shifts and preference flips under the current model. Samples with higher shortcut sensitivity are dynamically downweighted in the Bradley-Terry objective, encourag","authors_text":"Fei Sun, Fengyuan Liu, Mengnan Du, Yanguang Liu, Yongliang Miao, Zirui He","cross_cats":["cs.CL"],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-06-08T05:24:15Z","title":"DynaCF: Mitigating Shortcut Learning in Reward Models via Dynamic Counterfactual Sensitivity"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2606.09043","kind":"arxiv","version":1},"verdict":{"created_at":null,"id":null,"model_set":{},"one_line_summary":"","pipeline_version":null,"pith_extraction_headline":"","strongest_claim":"","weakest_assumption":""}},"verdict_id":null}}],"author_attestations":[],"timestamp_anchors":[],"storage_attestations":[],"citation_signatures":[],"replication_records":[],"corrections":[],"mirror_hints":[],"record_created":{"event_id":"sha256:d6d7f20b582b1a09b8b24484457e75af7345c02cb8546d504ddbc35bf02be2dd","target":"record","created_at":"2026-06-09T02:07:55Z","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":"3a549ff70ec4e7237cf6f3cc9b1fc4bc3eca7d7c175da07fc01b0727f76d877c","cross_cats_sorted":["cs.CL"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-06-08T05:24:15Z","title_canon_sha256":"bd8e00f4557352d50810b826a5d2203c756c40d3cab08c77a8c93d8ab66c0c69"},"schema_version":"1.0","source":{"id":"2606.09043","kind":"arxiv","version":1}},"canonical_sha256":"84e10cd89192d27d251985f26dcc280510705eb9bae68aee78ddd4279ae74c5f","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"84e10cd89192d27d251985f26dcc280510705eb9bae68aee78ddd4279ae74c5f","first_computed_at":"2026-06-09T02:07:55.294872Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-06-09T02:07:55.294872Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"xzMW1Fm4j01IxIk8WPS7+pxDQfB2W58bWXylZJVNN6wlrU5QsL4f23bPaD1RqjkrLvXnfVfP3n6GU4nm/Mt1Dw==","signature_status":"signed_v1","signed_at":"2026-06-09T02:07:55.295763Z","signed_message":"canonical_sha256_bytes"},"source_id":"2606.09043","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:d6d7f20b582b1a09b8b24484457e75af7345c02cb8546d504ddbc35bf02be2dd","sha256:7a36a5ba3a5762d1821ad0e47fae7e69c3d4a8c4543f49eeac7e5ae07446918a"],"state_sha256":"91406d17ef2d76656c29eaa4e3cb9051078c1aaaf445e5a5825a82da485a2cd9"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"T/D9aRHHmv/eWC7wPOQkAtkAMzYMzsgBCDIM7QRnRCHhSgFhTG/aj2wx6/KIGGk3tUd08BVKnjbKLNhZaPkIAw==","signed_message":"bundle_sha256_bytes","signed_at":"2026-06-30T00:07:31.026344Z","bundle_sha256":"36e07011e1c3746ca8cd8c0ad1678d06fdf709985cef6d82df6a8231e2c52ee2"}}