{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2026:G6BON2HG4JT5WVEOWT3BNBGVOL","short_pith_number":"pith:G6BON2HG","canonical_record":{"source":{"id":"2604.15215","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.RO","submitted_at":"2026-04-16T16:47:08Z","cross_cats_sorted":[],"title_canon_sha256":"8919fa8822be0ee0b38442bb96c1d16f44170cc3affe78f6fffbbc347ebc25be","abstract_canon_sha256":"ff645d7c776c74a41a146573031129e0bf487376c93a0ec3798471549a0f975c"},"schema_version":"1.0"},"canonical_sha256":"3782e6e8e6e267db548eb4f61684d572ec8e532c8275b6f134ad4147820c17ef","source":{"kind":"arxiv","id":"2604.15215","version":2},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2604.15215","created_at":"2026-05-20T00:00:38Z"},{"alias_kind":"arxiv_version","alias_value":"2604.15215v2","created_at":"2026-05-20T00:00:38Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2604.15215","created_at":"2026-05-20T00:00:38Z"},{"alias_kind":"pith_short_12","alias_value":"G6BON2HG4JT5","created_at":"2026-05-20T00:00:38Z"},{"alias_kind":"pith_short_16","alias_value":"G6BON2HG4JT5WVEO","created_at":"2026-05-20T00:00:38Z"},{"alias_kind":"pith_short_8","alias_value":"G6BON2HG","created_at":"2026-05-20T00:00:38Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2026:G6BON2HG4JT5WVEOWT3BNBGVOL","target":"record","payload":{"canonical_record":{"source":{"id":"2604.15215","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.RO","submitted_at":"2026-04-16T16:47:08Z","cross_cats_sorted":[],"title_canon_sha256":"8919fa8822be0ee0b38442bb96c1d16f44170cc3affe78f6fffbbc347ebc25be","abstract_canon_sha256":"ff645d7c776c74a41a146573031129e0bf487376c93a0ec3798471549a0f975c"},"schema_version":"1.0"},"canonical_sha256":"3782e6e8e6e267db548eb4f61684d572ec8e532c8275b6f134ad4147820c17ef","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-20T00:00:38.138523Z","signature_b64":"7kWHFYPY7YzRqAskVHHQUEXIXqycv8+2dH1I6rMdTDFZemQ7qT/nsVPaly3Tiblky1iaO3YpwPJLR06OSsZLAA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"3782e6e8e6e267db548eb4f61684d572ec8e532c8275b6f134ad4147820c17ef","last_reissued_at":"2026-05-20T00:00:38.137942Z","signature_status":"signed_v1","first_computed_at":"2026-05-20T00:00:38.137942Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2604.15215","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":"2026-05-20T00:00:38Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"z7o+XmOgKLRxPSg7WFSAqJscHpL5L0WGA/FZWnyq5tc3boZ3UApGbZBybznmLWM5Caldj56wVrQ9EUY7rc2MBA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-06-04T22:45:33.844892Z"},"content_sha256":"bc32ecce8f2ac64b0a4208d44ad384e5352e8e3b2e7c55f926d8650ed251319a","schema_version":"1.0","event_id":"sha256:bc32ecce8f2ac64b0a4208d44ad384e5352e8e3b2e7c55f926d8650ed251319a"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2026:G6BON2HG4JT5WVEOWT3BNBGVOL","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"A Hierarchical Spatiotemporal Action Tokenizer for In-Context Imitation Learning in Robotics","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"A two-level vector quantizer that clusters robot actions while also reconstructing their timestamps improves in-context imitation learning.","cross_cats":[],"primary_cat":"cs.RO","authors_text":"Ali Shah Ali, Andrey Konin, Fawad Javed Fateh, Murad Popattia, M. Zeeshan Zia, Quoc-Huy Tran, Usman Nizamani","submitted_at":"2026-04-16T16:47:08Z","abstract_excerpt":"We present a novel hierarchical spatiotemporal action tokenizer for in-context imitation learning. We first propose a hierarchical approach, which consists of two successive levels of vector quantization. In particular, the lower level assigns input actions to fine-grained subclusters, while the higher level further maps fine-grained subclusters to clusters. Our hierarchical approach outperforms the non-hierarchical counterpart, while mainly exploiting spatial information by reconstructing input actions. Furthermore, we extend our approach by utilizing both spatial and temporal cues, forming a"},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"Extensive evaluations on multiple simulation and real robotic manipulation benchmarks show that our approach establishes a new state-of-the-art performance in in-context imitation learning.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"The assumption that the hierarchical spatiotemporal clustering and joint reconstruction of actions plus timestamps will produce tokens that generalize beyond the specific benchmarks and robot platforms used in the evaluations.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"A two-level vector quantization tokenizer that jointly reconstructs robot actions and timestamps to improve in-context imitation learning performance on manipulation benchmarks.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"A two-level vector quantizer that clusters robot actions while also reconstructing their timestamps improves in-context imitation learning.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"facd1aa0e5d8555b11319ab38dc526825097a692ba02164df460fddbc075c352"},"source":{"id":"2604.15215","kind":"arxiv","version":2},"verdict":{"id":"5e7c0969-ddef-4b02-9f30-3539fd8e1081","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-19T17:20:03.419167Z","strongest_claim":"Extensive evaluations on multiple simulation and real robotic manipulation benchmarks show that our approach establishes a new state-of-the-art performance in in-context imitation learning.","one_line_summary":"A two-level vector quantization tokenizer that jointly reconstructs robot actions and timestamps to improve in-context imitation learning performance on manipulation benchmarks.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"The assumption that the hierarchical spatiotemporal clustering and joint reconstruction of actions plus timestamps will produce tokens that generalize beyond the specific benchmarks and robot platforms used in the evaluations.","pith_extraction_headline":"A two-level vector quantizer that clusters robot actions while also reconstructing their timestamps improves in-context imitation learning."},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2604.15215/integrity.json","findings":[],"available":true,"detectors_run":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938"},"references":{"count":52,"sample":[{"doi":"","year":2021,"title":"A. Mandlekar, D. Xu, J. Wong, S. Nasiriany, C. Wang, R. Kulkarni, L. Fei-Fei, S. Savarese, Y . Zhu, and R. Mart´ın-Mart´ın. What matters in learning from offline human demonstrations for robot manipul","work_id":"70db99d4-b7bf-4ca3-8395-8c071b286702","ref_index":1,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2024,"title":"Octo: An Open-Source Generalist Robot Policy","work_id":"f9ca0722-8855-48c3-a27a-0eefb7e19253","ref_index":2,"cited_arxiv_id":"2405.12213","is_internal_anchor":true},{"doi":"","year":2024,"title":"Open x-embodiment: Robotic learning datasets and rt-x models: Open x- embodiment collaboration 0","work_id":"c8fe1d83-8e88-41b8-ad0b-feb61a748902","ref_index":3,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2024,"title":"Khazatsky et al","work_id":"63a82a94-47f5-4f81-803e-9ec693af5607","ref_index":4,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2020,"title":"T. Brown, B. Mann, N. Ryder, M. Subbiah, J. D. Kaplan, P. Dhariwal, A. Neelakantan, P. Shyam, G. Sastry, A. Askell, et al. Language models are few-shot learners.NeurIPS, 2020","work_id":"c3601e58-863b-4181-a8b4-c4f8474aab09","ref_index":5,"cited_arxiv_id":"","is_internal_anchor":false}],"resolved_work":52,"snapshot_sha256":"c2241be740254df9ac51dbe81b68079d3c74542c081dfe96ff4e162c5002d2dd","internal_anchors":4},"formal_canon":{"evidence_count":2,"snapshot_sha256":"b1a6ef2c4b0222dbc2df4a792d1e447bc64b4414a7679bb7d4d9bce29ad374f1"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"},"verdict_id":"5e7c0969-ddef-4b02-9f30-3539fd8e1081"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-05-20T00:00:38Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"mcHr+o2Ur/AJ4Zmvni2vFerZe3HwIvwXKP9NVZ6TqV1n+AthaNH6C9nZj/XM0ut4gzsOy465ixBeMTyJQw7yAg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-06-04T22:45:33.846133Z"},"content_sha256":"b5a39f1652b3079c5b13436f21fbd5dd9b2078b2dc4a39462935e4816af00371","schema_version":"1.0","event_id":"sha256:b5a39f1652b3079c5b13436f21fbd5dd9b2078b2dc4a39462935e4816af00371"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/G6BON2HG4JT5WVEOWT3BNBGVOL/bundle.json","state_url":"https://pith.science/pith/G6BON2HG4JT5WVEOWT3BNBGVOL/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/G6BON2HG4JT5WVEOWT3BNBGVOL/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":"2026-06-04T22:45:33Z","links":{"resolver":"https://pith.science/pith/G6BON2HG4JT5WVEOWT3BNBGVOL","bundle":"https://pith.science/pith/G6BON2HG4JT5WVEOWT3BNBGVOL/bundle.json","state":"https://pith.science/pith/G6BON2HG4JT5WVEOWT3BNBGVOL/state.json","well_known_bundle":"https://pith.science/.well-known/pith/G6BON2HG4JT5WVEOWT3BNBGVOL/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:G6BON2HG4JT5WVEOWT3BNBGVOL","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":"ff645d7c776c74a41a146573031129e0bf487376c93a0ec3798471549a0f975c","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.RO","submitted_at":"2026-04-16T16:47:08Z","title_canon_sha256":"8919fa8822be0ee0b38442bb96c1d16f44170cc3affe78f6fffbbc347ebc25be"},"schema_version":"1.0","source":{"id":"2604.15215","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2604.15215","created_at":"2026-05-20T00:00:38Z"},{"alias_kind":"arxiv_version","alias_value":"2604.15215v2","created_at":"2026-05-20T00:00:38Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2604.15215","created_at":"2026-05-20T00:00:38Z"},{"alias_kind":"pith_short_12","alias_value":"G6BON2HG4JT5","created_at":"2026-05-20T00:00:38Z"},{"alias_kind":"pith_short_16","alias_value":"G6BON2HG4JT5WVEO","created_at":"2026-05-20T00:00:38Z"},{"alias_kind":"pith_short_8","alias_value":"G6BON2HG","created_at":"2026-05-20T00:00:38Z"}],"graph_snapshots":[{"event_id":"sha256:b5a39f1652b3079c5b13436f21fbd5dd9b2078b2dc4a39462935e4816af00371","target":"graph","created_at":"2026-05-20T00:00:38Z","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":"Extensive evaluations on multiple simulation and real robotic manipulation benchmarks show that our approach establishes a new state-of-the-art performance in in-context imitation learning."},{"attestation":"unclaimed","claim_id":"C2","kind":"weakest_assumption","source":"verdict.weakest_assumption","status":"machine_extracted","text":"The assumption that the hierarchical spatiotemporal clustering and joint reconstruction of actions plus timestamps will produce tokens that generalize beyond the specific benchmarks and robot platforms used in the evaluations."},{"attestation":"unclaimed","claim_id":"C3","kind":"one_line_summary","source":"verdict.one_line_summary","status":"machine_extracted","text":"A two-level vector quantization tokenizer that jointly reconstructs robot actions and timestamps to improve in-context imitation learning performance on manipulation benchmarks."},{"attestation":"unclaimed","claim_id":"C4","kind":"headline","source":"verdict.pith_extraction.headline","status":"machine_extracted","text":"A two-level vector quantizer that clusters robot actions while also reconstructing their timestamps improves in-context imitation learning."}],"snapshot_sha256":"facd1aa0e5d8555b11319ab38dc526825097a692ba02164df460fddbc075c352"},"formal_canon":{"evidence_count":2,"snapshot_sha256":"b1a6ef2c4b0222dbc2df4a792d1e447bc64b4414a7679bb7d4d9bce29ad374f1"},"integrity":{"available":true,"clean":true,"detectors_run":[],"endpoint":"/pith/2604.15215/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"We present a novel hierarchical spatiotemporal action tokenizer for in-context imitation learning. We first propose a hierarchical approach, which consists of two successive levels of vector quantization. In particular, the lower level assigns input actions to fine-grained subclusters, while the higher level further maps fine-grained subclusters to clusters. Our hierarchical approach outperforms the non-hierarchical counterpart, while mainly exploiting spatial information by reconstructing input actions. Furthermore, we extend our approach by utilizing both spatial and temporal cues, forming a","authors_text":"Ali Shah Ali, Andrey Konin, Fawad Javed Fateh, Murad Popattia, M. Zeeshan Zia, Quoc-Huy Tran, Usman Nizamani","cross_cats":[],"headline":"A two-level vector quantizer that clusters robot actions while also reconstructing their timestamps improves in-context imitation learning.","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.RO","submitted_at":"2026-04-16T16:47:08Z","title":"A Hierarchical Spatiotemporal Action Tokenizer for In-Context Imitation Learning in Robotics"},"references":{"count":52,"internal_anchors":4,"resolved_work":52,"sample":[{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":1,"title":"A. Mandlekar, D. Xu, J. Wong, S. Nasiriany, C. Wang, R. Kulkarni, L. Fei-Fei, S. Savarese, Y . Zhu, and R. Mart´ın-Mart´ın. What matters in learning from offline human demonstrations for robot manipul","work_id":"70db99d4-b7bf-4ca3-8395-8c071b286702","year":2021},{"cited_arxiv_id":"2405.12213","doi":"","is_internal_anchor":true,"ref_index":2,"title":"Octo: An Open-Source Generalist Robot Policy","work_id":"f9ca0722-8855-48c3-a27a-0eefb7e19253","year":2024},{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":3,"title":"Open x-embodiment: Robotic learning datasets and rt-x models: Open x- embodiment collaboration 0","work_id":"c8fe1d83-8e88-41b8-ad0b-feb61a748902","year":2024},{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":4,"title":"Khazatsky et al","work_id":"63a82a94-47f5-4f81-803e-9ec693af5607","year":2024},{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":5,"title":"T. Brown, B. Mann, N. Ryder, M. Subbiah, J. D. Kaplan, P. Dhariwal, A. Neelakantan, P. Shyam, G. Sastry, A. Askell, et al. Language models are few-shot learners.NeurIPS, 2020","work_id":"c3601e58-863b-4181-a8b4-c4f8474aab09","year":2020}],"snapshot_sha256":"c2241be740254df9ac51dbe81b68079d3c74542c081dfe96ff4e162c5002d2dd"},"source":{"id":"2604.15215","kind":"arxiv","version":2},"verdict":{"created_at":"2026-05-19T17:20:03.419167Z","id":"5e7c0969-ddef-4b02-9f30-3539fd8e1081","model_set":{"reader":"grok-4.3"},"one_line_summary":"A two-level vector quantization tokenizer that jointly reconstructs robot actions and timestamps to improve in-context imitation learning performance on manipulation benchmarks.","pipeline_version":"pith-pipeline@v0.9.0","pith_extraction_headline":"A two-level vector quantizer that clusters robot actions while also reconstructing their timestamps improves in-context imitation learning.","strongest_claim":"Extensive evaluations on multiple simulation and real robotic manipulation benchmarks show that our approach establishes a new state-of-the-art performance in in-context imitation learning.","weakest_assumption":"The assumption that the hierarchical spatiotemporal clustering and joint reconstruction of actions plus timestamps will produce tokens that generalize beyond the specific benchmarks and robot platforms used in the evaluations."}},"verdict_id":"5e7c0969-ddef-4b02-9f30-3539fd8e1081"}}],"author_attestations":[],"timestamp_anchors":[],"storage_attestations":[],"citation_signatures":[],"replication_records":[],"corrections":[],"mirror_hints":[],"record_created":{"event_id":"sha256:bc32ecce8f2ac64b0a4208d44ad384e5352e8e3b2e7c55f926d8650ed251319a","target":"record","created_at":"2026-05-20T00:00:38Z","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":"ff645d7c776c74a41a146573031129e0bf487376c93a0ec3798471549a0f975c","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.RO","submitted_at":"2026-04-16T16:47:08Z","title_canon_sha256":"8919fa8822be0ee0b38442bb96c1d16f44170cc3affe78f6fffbbc347ebc25be"},"schema_version":"1.0","source":{"id":"2604.15215","kind":"arxiv","version":2}},"canonical_sha256":"3782e6e8e6e267db548eb4f61684d572ec8e532c8275b6f134ad4147820c17ef","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"3782e6e8e6e267db548eb4f61684d572ec8e532c8275b6f134ad4147820c17ef","first_computed_at":"2026-05-20T00:00:38.137942Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-05-20T00:00:38.137942Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"7kWHFYPY7YzRqAskVHHQUEXIXqycv8+2dH1I6rMdTDFZemQ7qT/nsVPaly3Tiblky1iaO3YpwPJLR06OSsZLAA==","signature_status":"signed_v1","signed_at":"2026-05-20T00:00:38.138523Z","signed_message":"canonical_sha256_bytes"},"source_id":"2604.15215","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:bc32ecce8f2ac64b0a4208d44ad384e5352e8e3b2e7c55f926d8650ed251319a","sha256:b5a39f1652b3079c5b13436f21fbd5dd9b2078b2dc4a39462935e4816af00371"],"state_sha256":"1ebb305bbfdc98b57e7ea2b5f4f166dde00c1e00a05b557a4702e230601fd7c6"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"yZmwvWTu/RtSjUN6xQSXrmxwxV84rHqv1cb2rHVY+tStqZqQPKw0JFx8Yd2gt8BU3wt8WXiBI3LTmZu2+WyqAw==","signed_message":"bundle_sha256_bytes","signed_at":"2026-06-04T22:45:33.850856Z","bundle_sha256":"59a68c4e656455a25af8c1399455f9fafa107b262297293b66724ac89beda7e0"}}