{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2026:DGX3YQPZT5L2JQOQSSFQIZ7XXQ","short_pith_number":"pith:DGX3YQPZ","canonical_record":{"source":{"id":"2605.21661","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.LG","submitted_at":"2026-05-20T19:13:28Z","cross_cats_sorted":["cs.AI","cs.CV"],"title_canon_sha256":"fcc4e2956ede4a0253a6a95365636381c8ebcec06bc51660398a3fac48b30a07","abstract_canon_sha256":"565271f404e36630664efbdcbfb417ec3287f2136a6e5af8328b509d6c8ad536"},"schema_version":"1.0"},"canonical_sha256":"19afbc41f99f57a4c1d0948b0467f7bc138183fb46a5c2e5479fe7afe5a8345a","source":{"kind":"arxiv","id":"2605.21661","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2605.21661","created_at":"2026-05-22T01:03:27Z"},{"alias_kind":"arxiv_version","alias_value":"2605.21661v1","created_at":"2026-05-22T01:03:27Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.21661","created_at":"2026-05-22T01:03:27Z"},{"alias_kind":"pith_short_12","alias_value":"DGX3YQPZT5L2","created_at":"2026-05-22T01:03:27Z"},{"alias_kind":"pith_short_16","alias_value":"DGX3YQPZT5L2JQOQ","created_at":"2026-05-22T01:03:27Z"},{"alias_kind":"pith_short_8","alias_value":"DGX3YQPZ","created_at":"2026-05-22T01:03:27Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2026:DGX3YQPZT5L2JQOQSSFQIZ7XXQ","target":"record","payload":{"canonical_record":{"source":{"id":"2605.21661","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.LG","submitted_at":"2026-05-20T19:13:28Z","cross_cats_sorted":["cs.AI","cs.CV"],"title_canon_sha256":"fcc4e2956ede4a0253a6a95365636381c8ebcec06bc51660398a3fac48b30a07","abstract_canon_sha256":"565271f404e36630664efbdcbfb417ec3287f2136a6e5af8328b509d6c8ad536"},"schema_version":"1.0"},"canonical_sha256":"19afbc41f99f57a4c1d0948b0467f7bc138183fb46a5c2e5479fe7afe5a8345a","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-22T01:03:27.160604Z","signature_b64":"4iGD5ObwbqPw+U/QyFSHcwO7hhlJgqW/eIurziTl7sJ/s7GRVTCloAXCHpxM0JK55RtG4OLKs6JZGgZjG9KjAw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"19afbc41f99f57a4c1d0948b0467f7bc138183fb46a5c2e5479fe7afe5a8345a","last_reissued_at":"2026-05-22T01:03:27.160155Z","signature_status":"signed_v1","first_computed_at":"2026-05-22T01:03:27.160155Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2605.21661","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":"2026-05-22T01:03:27Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"wl8Yr93TBLWkal7L3u0LFVPDaI+ZYQPyAV7zrTVObT6ABmiILmyuGYDMCfJaAcgRlj2P008u38a2WO7ncxDBCA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-06-05T16:32:03.606465Z"},"content_sha256":"9d2cf90647c9de07ab115aade08d6d3ff1688196cab9625e9b9aacf4ab2cf3f2","schema_version":"1.0","event_id":"sha256:9d2cf90647c9de07ab115aade08d6d3ff1688196cab9625e9b9aacf4ab2cf3f2"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2026:DGX3YQPZT5L2JQOQSSFQIZ7XXQ","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Hierarchical Variational Policies for Reward-Guided Diffusion","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","headline":"","cross_cats":["cs.AI","cs.CV"],"primary_cat":"cs.LG","authors_text":"Farrin Marouf Sofian, Felix Draxler, Jan Niklas Groeneveld, Kushagra Pandey, Stephan Mandt","submitted_at":"2026-05-20T19:13:28Z","abstract_excerpt":"Adapting pretrained diffusion models to downstream objectives such as inverse problems often requires expensive test-time guidance or optimization. We propose a principled framework for generating high-quality reward-aligned samples at substantially reduced inference cost. Our approach formulates test-time adaptation as a hierarchical variational model, where control is amortized into a lightweight yet expressive stochastic policy. This formulation naturally supports few-step diffusion sampling: large step sizes enable fast inference, while the learned policy maintains sample quality by provid"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2605.21661","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/2605.21661/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":"2026-05-22T01:03:27Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"+Ecbm+XxJcZe6boSCL9OTcNgpVT2u+KzZ8JC5zZ+CQsO7WRPufBjWt7G1FYT8ucywcyaDxczAu3C7ZaAWN2sBA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-06-05T16:32:03.606847Z"},"content_sha256":"f1ad3b2e42c756433127f6f9700b4af1920d1905073f19fb0e8960c7088584c1","schema_version":"1.0","event_id":"sha256:f1ad3b2e42c756433127f6f9700b4af1920d1905073f19fb0e8960c7088584c1"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/DGX3YQPZT5L2JQOQSSFQIZ7XXQ/bundle.json","state_url":"https://pith.science/pith/DGX3YQPZT5L2JQOQSSFQIZ7XXQ/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/DGX3YQPZT5L2JQOQSSFQIZ7XXQ/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":"2026-06-05T16:32:03Z","links":{"resolver":"https://pith.science/pith/DGX3YQPZT5L2JQOQSSFQIZ7XXQ","bundle":"https://pith.science/pith/DGX3YQPZT5L2JQOQSSFQIZ7XXQ/bundle.json","state":"https://pith.science/pith/DGX3YQPZT5L2JQOQSSFQIZ7XXQ/state.json","well_known_bundle":"https://pith.science/.well-known/pith/DGX3YQPZT5L2JQOQSSFQIZ7XXQ/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:DGX3YQPZT5L2JQOQSSFQIZ7XXQ","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":"565271f404e36630664efbdcbfb417ec3287f2136a6e5af8328b509d6c8ad536","cross_cats_sorted":["cs.AI","cs.CV"],"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.LG","submitted_at":"2026-05-20T19:13:28Z","title_canon_sha256":"fcc4e2956ede4a0253a6a95365636381c8ebcec06bc51660398a3fac48b30a07"},"schema_version":"1.0","source":{"id":"2605.21661","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2605.21661","created_at":"2026-05-22T01:03:27Z"},{"alias_kind":"arxiv_version","alias_value":"2605.21661v1","created_at":"2026-05-22T01:03:27Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.21661","created_at":"2026-05-22T01:03:27Z"},{"alias_kind":"pith_short_12","alias_value":"DGX3YQPZT5L2","created_at":"2026-05-22T01:03:27Z"},{"alias_kind":"pith_short_16","alias_value":"DGX3YQPZT5L2JQOQ","created_at":"2026-05-22T01:03:27Z"},{"alias_kind":"pith_short_8","alias_value":"DGX3YQPZ","created_at":"2026-05-22T01:03:27Z"}],"graph_snapshots":[{"event_id":"sha256:f1ad3b2e42c756433127f6f9700b4af1920d1905073f19fb0e8960c7088584c1","target":"graph","created_at":"2026-05-22T01:03:27Z","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/2605.21661/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Adapting pretrained diffusion models to downstream objectives such as inverse problems often requires expensive test-time guidance or optimization. We propose a principled framework for generating high-quality reward-aligned samples at substantially reduced inference cost. Our approach formulates test-time adaptation as a hierarchical variational model, where control is amortized into a lightweight yet expressive stochastic policy. This formulation naturally supports few-step diffusion sampling: large step sizes enable fast inference, while the learned policy maintains sample quality by provid","authors_text":"Farrin Marouf Sofian, Felix Draxler, Jan Niklas Groeneveld, Kushagra Pandey, Stephan Mandt","cross_cats":["cs.AI","cs.CV"],"headline":"","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.LG","submitted_at":"2026-05-20T19:13:28Z","title":"Hierarchical Variational Policies for Reward-Guided Diffusion"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2605.21661","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":"sha256:9d2cf90647c9de07ab115aade08d6d3ff1688196cab9625e9b9aacf4ab2cf3f2","target":"record","created_at":"2026-05-22T01:03:27Z","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":"565271f404e36630664efbdcbfb417ec3287f2136a6e5af8328b509d6c8ad536","cross_cats_sorted":["cs.AI","cs.CV"],"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.LG","submitted_at":"2026-05-20T19:13:28Z","title_canon_sha256":"fcc4e2956ede4a0253a6a95365636381c8ebcec06bc51660398a3fac48b30a07"},"schema_version":"1.0","source":{"id":"2605.21661","kind":"arxiv","version":1}},"canonical_sha256":"19afbc41f99f57a4c1d0948b0467f7bc138183fb46a5c2e5479fe7afe5a8345a","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"19afbc41f99f57a4c1d0948b0467f7bc138183fb46a5c2e5479fe7afe5a8345a","first_computed_at":"2026-05-22T01:03:27.160155Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-05-22T01:03:27.160155Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"4iGD5ObwbqPw+U/QyFSHcwO7hhlJgqW/eIurziTl7sJ/s7GRVTCloAXCHpxM0JK55RtG4OLKs6JZGgZjG9KjAw==","signature_status":"signed_v1","signed_at":"2026-05-22T01:03:27.160604Z","signed_message":"canonical_sha256_bytes"},"source_id":"2605.21661","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:9d2cf90647c9de07ab115aade08d6d3ff1688196cab9625e9b9aacf4ab2cf3f2","sha256:f1ad3b2e42c756433127f6f9700b4af1920d1905073f19fb0e8960c7088584c1"],"state_sha256":"16d46fb1b95c43324d1d4c2137654df958538b890215636e4431b25807a3194a"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"qsryzi085QlizKhfN8IPnJP16ni1gUrQmWlsaliVtFjoDDk1q0Vri+ADoV+tG6PFLhBihsIOLADNXY+0hHMtDA==","signed_message":"bundle_sha256_bytes","signed_at":"2026-06-05T16:32:03.608797Z","bundle_sha256":"4cf1c3f57276b63b75f15fc83b417883575109c2f235ba2a3078065b7d07125a"}}