{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2024:MS46TU7AT7NWFJXSCCLLT55RPY","short_pith_number":"pith:MS46TU7A","canonical_record":{"source":{"id":"2408.08812","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2024-08-16T15:47:08Z","cross_cats_sorted":[],"title_canon_sha256":"74c38ce0cc5ba376e718df0efd304a483af9a326f4ad7ee62c89e829b5399e87","abstract_canon_sha256":"8396732ad5487868120f5b0c2cc45140a2d84f2857739732d46b89926ada2d34"},"schema_version":"1.0"},"canonical_sha256":"64b9e9d3e09fdb62a6f21096b9f7b17e16a7827551878c06408ff8ca0b749524","source":{"kind":"arxiv","id":"2408.08812","version":2},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2408.08812","created_at":"2026-05-21T01:04:11Z"},{"alias_kind":"arxiv_version","alias_value":"2408.08812v2","created_at":"2026-05-21T01:04:11Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2408.08812","created_at":"2026-05-21T01:04:11Z"},{"alias_kind":"pith_short_12","alias_value":"MS46TU7AT7NW","created_at":"2026-05-21T01:04:11Z"},{"alias_kind":"pith_short_16","alias_value":"MS46TU7AT7NWFJXS","created_at":"2026-05-21T01:04:11Z"},{"alias_kind":"pith_short_8","alias_value":"MS46TU7A","created_at":"2026-05-21T01:04:11Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2024:MS46TU7AT7NWFJXSCCLLT55RPY","target":"record","payload":{"canonical_record":{"source":{"id":"2408.08812","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2024-08-16T15:47:08Z","cross_cats_sorted":[],"title_canon_sha256":"74c38ce0cc5ba376e718df0efd304a483af9a326f4ad7ee62c89e829b5399e87","abstract_canon_sha256":"8396732ad5487868120f5b0c2cc45140a2d84f2857739732d46b89926ada2d34"},"schema_version":"1.0"},"canonical_sha256":"64b9e9d3e09fdb62a6f21096b9f7b17e16a7827551878c06408ff8ca0b749524","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-21T01:04:11.934990Z","signature_b64":"NIe12LzW3NHF5D2VrvkrbLXnSrKrlQiXeoWJ462p0b3l2knX2THuj6XjS6cXAWbW5KK0P3Zx/ykhnvx8DMe9Dg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"64b9e9d3e09fdb62a6f21096b9f7b17e16a7827551878c06408ff8ca0b749524","last_reissued_at":"2026-05-21T01:04:11.934293Z","signature_status":"signed_v1","first_computed_at":"2026-05-21T01:04:11.934293Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2408.08812","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":"2026-05-21T01:04:11Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"HBQKq0/gh4f6hPR4qvypgLigfhM0Bl09PMsYTOy2Er8fCo/Imdg5jPhW9GAIqjlyVUZwc3nf4ldumujidhVCAw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-05-26T10:35:27.043908Z"},"content_sha256":"c7261a7179a481c8b971036b6108cd1411162dad53c30195f7a29e57b9b94ef4","schema_version":"1.0","event_id":"sha256:c7261a7179a481c8b971036b6108cd1411162dad53c30195f7a29e57b9b94ef4"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2024:MS46TU7AT7NWFJXSCCLLT55RPY","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"TRAM: Test-Time Risk Adaptation with Mixture of Agents","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Amrit Singh Bedi, Amy Zhang, Hao Zhu, Mohamad Fares El Hajj Chehade","submitted_at":"2024-08-16T15:47:08Z","abstract_excerpt":"Deployed reinforcement learning agents often face safety requirements that are specified only after training, such as new hazard maps, revised risk thresholds, or behavioral alignment constraints. We study zero-update deployment-time adaptation, where a fixed library of risk-neutral source policies is reused under a newly specified reward-risk tradeoff. We propose TRAM (Test-Time Risk Adaptation via Mixture of Agents), a source-scored composition rule that evaluates each source policy under the target reward and an occupancy-based deployment risk, then selects actions using risk-adjusted sourc"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2408.08812","kind":"arxiv","version":2},"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/2408.08812/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":"2026-05-21T01:04:11Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"EC0aLpf1sx29yLcAAChWu64kIgroL+r3gGoCginZdvfKzFhiB/IYE4FAgBv7m3UpRWPZNf8gv7kDiVJPRaQQBw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-05-26T10:35:27.044415Z"},"content_sha256":"7ce747d358091d2d61f0491e5e970e7f8be526073872fb395446ecb7cc7dd4f3","schema_version":"1.0","event_id":"sha256:7ce747d358091d2d61f0491e5e970e7f8be526073872fb395446ecb7cc7dd4f3"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/MS46TU7AT7NWFJXSCCLLT55RPY/bundle.json","state_url":"https://pith.science/pith/MS46TU7AT7NWFJXSCCLLT55RPY/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/MS46TU7AT7NWFJXSCCLLT55RPY/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":"2026-05-26T10:35:27Z","links":{"resolver":"https://pith.science/pith/MS46TU7AT7NWFJXSCCLLT55RPY","bundle":"https://pith.science/pith/MS46TU7AT7NWFJXSCCLLT55RPY/bundle.json","state":"https://pith.science/pith/MS46TU7AT7NWFJXSCCLLT55RPY/state.json","well_known_bundle":"https://pith.science/.well-known/pith/MS46TU7AT7NWFJXSCCLLT55RPY/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2024:MS46TU7AT7NWFJXSCCLLT55RPY","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":"8396732ad5487868120f5b0c2cc45140a2d84f2857739732d46b89926ada2d34","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2024-08-16T15:47:08Z","title_canon_sha256":"74c38ce0cc5ba376e718df0efd304a483af9a326f4ad7ee62c89e829b5399e87"},"schema_version":"1.0","source":{"id":"2408.08812","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2408.08812","created_at":"2026-05-21T01:04:11Z"},{"alias_kind":"arxiv_version","alias_value":"2408.08812v2","created_at":"2026-05-21T01:04:11Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2408.08812","created_at":"2026-05-21T01:04:11Z"},{"alias_kind":"pith_short_12","alias_value":"MS46TU7AT7NW","created_at":"2026-05-21T01:04:11Z"},{"alias_kind":"pith_short_16","alias_value":"MS46TU7AT7NWFJXS","created_at":"2026-05-21T01:04:11Z"},{"alias_kind":"pith_short_8","alias_value":"MS46TU7A","created_at":"2026-05-21T01:04:11Z"}],"graph_snapshots":[{"event_id":"sha256:7ce747d358091d2d61f0491e5e970e7f8be526073872fb395446ecb7cc7dd4f3","target":"graph","created_at":"2026-05-21T01:04:11Z","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/2408.08812/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Deployed reinforcement learning agents often face safety requirements that are specified only after training, such as new hazard maps, revised risk thresholds, or behavioral alignment constraints. We study zero-update deployment-time adaptation, where a fixed library of risk-neutral source policies is reused under a newly specified reward-risk tradeoff. We propose TRAM (Test-Time Risk Adaptation via Mixture of Agents), a source-scored composition rule that evaluates each source policy under the target reward and an occupancy-based deployment risk, then selects actions using risk-adjusted sourc","authors_text":"Amrit Singh Bedi, Amy Zhang, Hao Zhu, Mohamad Fares El Hajj Chehade","cross_cats":[],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2024-08-16T15:47:08Z","title":"TRAM: Test-Time Risk Adaptation with Mixture of Agents"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2408.08812","kind":"arxiv","version":2},"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:c7261a7179a481c8b971036b6108cd1411162dad53c30195f7a29e57b9b94ef4","target":"record","created_at":"2026-05-21T01:04:11Z","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":"8396732ad5487868120f5b0c2cc45140a2d84f2857739732d46b89926ada2d34","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2024-08-16T15:47:08Z","title_canon_sha256":"74c38ce0cc5ba376e718df0efd304a483af9a326f4ad7ee62c89e829b5399e87"},"schema_version":"1.0","source":{"id":"2408.08812","kind":"arxiv","version":2}},"canonical_sha256":"64b9e9d3e09fdb62a6f21096b9f7b17e16a7827551878c06408ff8ca0b749524","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"64b9e9d3e09fdb62a6f21096b9f7b17e16a7827551878c06408ff8ca0b749524","first_computed_at":"2026-05-21T01:04:11.934293Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-05-21T01:04:11.934293Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"NIe12LzW3NHF5D2VrvkrbLXnSrKrlQiXeoWJ462p0b3l2knX2THuj6XjS6cXAWbW5KK0P3Zx/ykhnvx8DMe9Dg==","signature_status":"signed_v1","signed_at":"2026-05-21T01:04:11.934990Z","signed_message":"canonical_sha256_bytes"},"source_id":"2408.08812","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:c7261a7179a481c8b971036b6108cd1411162dad53c30195f7a29e57b9b94ef4","sha256:7ce747d358091d2d61f0491e5e970e7f8be526073872fb395446ecb7cc7dd4f3"],"state_sha256":"e95e69c7b48db2b717426828170ef0a07f42a33847c0f8e64a92ba76306b216a"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"1UyHJLtM6HtkchPKEIKVApsk2GTeE1hokN1g2PlKzLWILuvrB48xLBToXwfe4zz0IuJii/vX2VjkSo65LrSfBA==","signed_message":"bundle_sha256_bytes","signed_at":"2026-05-26T10:35:27.047517Z","bundle_sha256":"27d3ab7ca0f8b41a273a92cb6b3b7795f9bce970246fada441bafbfd4272f5bc"}}