{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2026:FXSPYSGK4PSBLI4G62QLT5VJJV","short_pith_number":"pith:FXSPYSGK","canonical_record":{"source":{"id":"2606.09701","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2026-06-08T16:21:36Z","cross_cats_sorted":["cs.AI","cs.LG"],"title_canon_sha256":"335605a47ccb43370886ca71509a5347e0b22bb3be3413fc87ed97d1b1a15a64","abstract_canon_sha256":"a5d6e8c5ac05ddedb6f684fb2ef5e57bf138762b969913d36d8758b910be7e69"},"schema_version":"1.0"},"canonical_sha256":"2de4fc48cae3e415a386f6a0b9f6a94d73d11d1a57656bee8b92849b00656c77","source":{"kind":"arxiv","id":"2606.09701","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2606.09701","created_at":"2026-06-09T02:09:04Z"},{"alias_kind":"arxiv_version","alias_value":"2606.09701v1","created_at":"2026-06-09T02:09:04Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2606.09701","created_at":"2026-06-09T02:09:04Z"},{"alias_kind":"pith_short_12","alias_value":"FXSPYSGK4PSB","created_at":"2026-06-09T02:09:04Z"},{"alias_kind":"pith_short_16","alias_value":"FXSPYSGK4PSBLI4G","created_at":"2026-06-09T02:09:04Z"},{"alias_kind":"pith_short_8","alias_value":"FXSPYSGK","created_at":"2026-06-09T02:09:04Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2026:FXSPYSGK4PSBLI4G62QLT5VJJV","target":"record","payload":{"canonical_record":{"source":{"id":"2606.09701","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2026-06-08T16:21:36Z","cross_cats_sorted":["cs.AI","cs.LG"],"title_canon_sha256":"335605a47ccb43370886ca71509a5347e0b22bb3be3413fc87ed97d1b1a15a64","abstract_canon_sha256":"a5d6e8c5ac05ddedb6f684fb2ef5e57bf138762b969913d36d8758b910be7e69"},"schema_version":"1.0"},"canonical_sha256":"2de4fc48cae3e415a386f6a0b9f6a94d73d11d1a57656bee8b92849b00656c77","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-09T02:09:04.671779Z","signature_b64":"gGnuOU7csLpiqmtA6DTR8E0Rt3vBpxJlwlC1DdTZBE+lr4it2Myv6E8JkO2UcNvbxCL84eLNmi+nzZAyBV1fDQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"2de4fc48cae3e415a386f6a0b9f6a94d73d11d1a57656bee8b92849b00656c77","last_reissued_at":"2026-06-09T02:09:04.664889Z","signature_status":"signed_v1","first_computed_at":"2026-06-09T02:09:04.664889Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2606.09701","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":"2026-06-09T02:09:04Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"XqqujfR5w8ywc0ZcXO8IEEl/Ns0ajGiuILvB8UUh480BHdEjutZ2mPOsf3LxWxrvY9db5YB/rk9dJQzJR4oZBw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-06-23T06:39:21.155065Z"},"content_sha256":"63ff41da71ab50e1a7824dc81b62577d9685d9771983de196f9b8386cee33da9","schema_version":"1.0","event_id":"sha256:63ff41da71ab50e1a7824dc81b62577d9685d9771983de196f9b8386cee33da9"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2026:FXSPYSGK4PSBLI4G62QLT5VJJV","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Learning to Attack and Defend: Adaptive Red Teaming of Language Models via GRPO","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.LG"],"primary_cat":"cs.CL","authors_text":"Amanda Minnich, Blake Bullwinkel, Eugenia Kim, Mark Russinovich","submitted_at":"2026-06-08T16:21:36Z","abstract_excerpt":"AI red teaming must continually adapt to evolving attackers and defenders. Reinforcement learning offers a promising approach to discovering novel attacks, and co-training methods can produce more robust defenders in tandem. Recent works have demonstrated the efficacy of attacker-defender co-training by applying PPO and DPO, but report that GRPO is unstable in this setting. We introduce AdvGRPO, a co-training framework that makes GRPO viable for joint attacker-defender optimization using dense multi-channel rewards and decoupled advantage normalization. Training progresses through a curriculum"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2606.09701","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/2606.09701/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":"2026-06-09T02:09:04Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"jTL8Z8QfY+HKCBHQZq7T80q99o1qQwq4cf7sCP3HhGqX5/+g0MMTTfyONsr9Mt72aK21x6YBB9NLdAt8kj3qDg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-06-23T06:39:21.155454Z"},"content_sha256":"bf2033d7b110b4d9d28de98cf40af9653f0d058cdff2b22204c914824307a63f","schema_version":"1.0","event_id":"sha256:bf2033d7b110b4d9d28de98cf40af9653f0d058cdff2b22204c914824307a63f"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/FXSPYSGK4PSBLI4G62QLT5VJJV/bundle.json","state_url":"https://pith.science/pith/FXSPYSGK4PSBLI4G62QLT5VJJV/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/FXSPYSGK4PSBLI4G62QLT5VJJV/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":"2026-06-23T06:39:21Z","links":{"resolver":"https://pith.science/pith/FXSPYSGK4PSBLI4G62QLT5VJJV","bundle":"https://pith.science/pith/FXSPYSGK4PSBLI4G62QLT5VJJV/bundle.json","state":"https://pith.science/pith/FXSPYSGK4PSBLI4G62QLT5VJJV/state.json","well_known_bundle":"https://pith.science/.well-known/pith/FXSPYSGK4PSBLI4G62QLT5VJJV/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:FXSPYSGK4PSBLI4G62QLT5VJJV","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":"a5d6e8c5ac05ddedb6f684fb2ef5e57bf138762b969913d36d8758b910be7e69","cross_cats_sorted":["cs.AI","cs.LG"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2026-06-08T16:21:36Z","title_canon_sha256":"335605a47ccb43370886ca71509a5347e0b22bb3be3413fc87ed97d1b1a15a64"},"schema_version":"1.0","source":{"id":"2606.09701","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2606.09701","created_at":"2026-06-09T02:09:04Z"},{"alias_kind":"arxiv_version","alias_value":"2606.09701v1","created_at":"2026-06-09T02:09:04Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2606.09701","created_at":"2026-06-09T02:09:04Z"},{"alias_kind":"pith_short_12","alias_value":"FXSPYSGK4PSB","created_at":"2026-06-09T02:09:04Z"},{"alias_kind":"pith_short_16","alias_value":"FXSPYSGK4PSBLI4G","created_at":"2026-06-09T02:09:04Z"},{"alias_kind":"pith_short_8","alias_value":"FXSPYSGK","created_at":"2026-06-09T02:09:04Z"}],"graph_snapshots":[{"event_id":"sha256:bf2033d7b110b4d9d28de98cf40af9653f0d058cdff2b22204c914824307a63f","target":"graph","created_at":"2026-06-09T02:09:04Z","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/2606.09701/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"AI red teaming must continually adapt to evolving attackers and defenders. Reinforcement learning offers a promising approach to discovering novel attacks, and co-training methods can produce more robust defenders in tandem. Recent works have demonstrated the efficacy of attacker-defender co-training by applying PPO and DPO, but report that GRPO is unstable in this setting. We introduce AdvGRPO, a co-training framework that makes GRPO viable for joint attacker-defender optimization using dense multi-channel rewards and decoupled advantage normalization. Training progresses through a curriculum","authors_text":"Amanda Minnich, Blake Bullwinkel, Eugenia Kim, Mark Russinovich","cross_cats":["cs.AI","cs.LG"],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2026-06-08T16:21:36Z","title":"Learning to Attack and Defend: Adaptive Red Teaming of Language Models via GRPO"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2606.09701","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":"sha256:63ff41da71ab50e1a7824dc81b62577d9685d9771983de196f9b8386cee33da9","target":"record","created_at":"2026-06-09T02:09:04Z","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":"a5d6e8c5ac05ddedb6f684fb2ef5e57bf138762b969913d36d8758b910be7e69","cross_cats_sorted":["cs.AI","cs.LG"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2026-06-08T16:21:36Z","title_canon_sha256":"335605a47ccb43370886ca71509a5347e0b22bb3be3413fc87ed97d1b1a15a64"},"schema_version":"1.0","source":{"id":"2606.09701","kind":"arxiv","version":1}},"canonical_sha256":"2de4fc48cae3e415a386f6a0b9f6a94d73d11d1a57656bee8b92849b00656c77","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"2de4fc48cae3e415a386f6a0b9f6a94d73d11d1a57656bee8b92849b00656c77","first_computed_at":"2026-06-09T02:09:04.664889Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-06-09T02:09:04.664889Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"gGnuOU7csLpiqmtA6DTR8E0Rt3vBpxJlwlC1DdTZBE+lr4it2Myv6E8JkO2UcNvbxCL84eLNmi+nzZAyBV1fDQ==","signature_status":"signed_v1","signed_at":"2026-06-09T02:09:04.671779Z","signed_message":"canonical_sha256_bytes"},"source_id":"2606.09701","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:63ff41da71ab50e1a7824dc81b62577d9685d9771983de196f9b8386cee33da9","sha256:bf2033d7b110b4d9d28de98cf40af9653f0d058cdff2b22204c914824307a63f"],"state_sha256":"6b86bb9753384f9b162c9918116d4b86bccaf9b986d8e6a7ee3f3ec235fcd4d1"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"k4WOb3E62niy8s+78sPKDYgbFr5kxZgu/r8RN0OEVNQbZlr+D1Z7Tlyoo9jFfCpwA/wGcL1SniAnFQKo473KCg==","signed_message":"bundle_sha256_bytes","signed_at":"2026-06-23T06:39:21.157392Z","bundle_sha256":"38effb2c613ad17191b083a9f54923e85c7ed90d5518ebdf0e5c54d614b3ce72"}}