{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2024:K5H7VTRJZIBKMHCPQ3OPZ3ZJBH","short_pith_number":"pith:K5H7VTRJ","canonical_record":{"source":{"id":"2410.01926","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2024-10-02T18:20:51Z","cross_cats_sorted":[],"title_canon_sha256":"806f15cab1868d82365f85ee75f05f34a95e7db58a442b764d52b3e3482dfb30","abstract_canon_sha256":"449c4945f771fd25b4e87a4308208b5add04c4782e3127c6ac0e4db85317bc21"},"schema_version":"1.0"},"canonical_sha256":"574fface29ca02a61c4f86dcfcef2909e785000da4b9f7dbc792d248dfe25881","source":{"kind":"arxiv","id":"2410.01926","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2410.01926","created_at":"2026-07-05T09:14:57Z"},{"alias_kind":"arxiv_version","alias_value":"2410.01926v1","created_at":"2026-07-05T09:14:57Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2410.01926","created_at":"2026-07-05T09:14:57Z"},{"alias_kind":"pith_short_12","alias_value":"K5H7VTRJZIBK","created_at":"2026-07-05T09:14:57Z"},{"alias_kind":"pith_short_16","alias_value":"K5H7VTRJZIBKMHCP","created_at":"2026-07-05T09:14:57Z"},{"alias_kind":"pith_short_8","alias_value":"K5H7VTRJ","created_at":"2026-07-05T09:14:57Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2024:K5H7VTRJZIBKMHCPQ3OPZ3ZJBH","target":"record","payload":{"canonical_record":{"source":{"id":"2410.01926","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2024-10-02T18:20:51Z","cross_cats_sorted":[],"title_canon_sha256":"806f15cab1868d82365f85ee75f05f34a95e7db58a442b764d52b3e3482dfb30","abstract_canon_sha256":"449c4945f771fd25b4e87a4308208b5add04c4782e3127c6ac0e4db85317bc21"},"schema_version":"1.0"},"canonical_sha256":"574fface29ca02a61c4f86dcfcef2909e785000da4b9f7dbc792d248dfe25881","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T09:14:57.780820Z","signature_b64":"zVvCGS13Z0PqnN1/51exvZhQotn8WXQe4GOuq2xGeXLJmUugbUxj29e6DjJA5rVkR7HJFOKsUO+KhGWPhrgrAA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"574fface29ca02a61c4f86dcfcef2909e785000da4b9f7dbc792d248dfe25881","last_reissued_at":"2026-07-05T09:14:57.780408Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T09:14:57.780408Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2410.01926","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":"2026-07-05T09:14:57Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"JjGhhcExFsyZ4wzcvVJL8sd5c9ANygoupYHA6vHFiN4WLBiPtI9fny7YpLIdlI9GkE76Z62wXw24aU+Qrcg/DA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-07-07T07:10:13.618338Z"},"content_sha256":"5a814b45c8cca326578c3b09bd5b23f9f966c2e38e3ac157d273af4321cf63ae","schema_version":"1.0","event_id":"sha256:5a814b45c8cca326578c3b09bd5b23f9f966c2e38e3ac157d273af4321cf63ae"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2024:K5H7VTRJZIBKMHCPQ3OPZ3ZJBH","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"MARPLE: A Benchmark for Long-Horizon Inference","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Emily Jin, Erik Brockbank, Hannah Cha, Jan-Philipp Fr\\\"anken, Jiajun Wu, Ruohan Zhang, Sarah Wu, Tobias Gerstenberg, Weiyu Liu, Zhuoyi Huang","submitted_at":"2024-10-02T18:20:51Z","abstract_excerpt":"Reconstructing past events requires reasoning across long time horizons. To figure out what happened, we need to use our prior knowledge about the world and human behavior and draw inferences from various sources of evidence including visual, language, and auditory cues. We introduce MARPLE, a benchmark for evaluating long-horizon inference capabilities using multi-modal evidence. Our benchmark features agents interacting with simulated households, supporting vision, language, and auditory stimuli, as well as procedurally generated environments and agent behaviors. Inspired by classic ``whodun"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2410.01926","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/2410.01926/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":"2026-07-05T09:14:57Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"NRJ9qG1SUk71U6yRRYdz3cPK05qB+gI1os/P1sqUecM06e1/wf8KZbo1iM2NsCv3z9B0BT9Y0x8u42MA4XudDg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-07-07T07:10:13.618716Z"},"content_sha256":"51f585f269000f9ef7e15a5840f6b7b5557a485f2d11f972d0be3713b9c5cf14","schema_version":"1.0","event_id":"sha256:51f585f269000f9ef7e15a5840f6b7b5557a485f2d11f972d0be3713b9c5cf14"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/K5H7VTRJZIBKMHCPQ3OPZ3ZJBH/bundle.json","state_url":"https://pith.science/pith/K5H7VTRJZIBKMHCPQ3OPZ3ZJBH/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/K5H7VTRJZIBKMHCPQ3OPZ3ZJBH/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":"2026-07-07T07:10:13Z","links":{"resolver":"https://pith.science/pith/K5H7VTRJZIBKMHCPQ3OPZ3ZJBH","bundle":"https://pith.science/pith/K5H7VTRJZIBKMHCPQ3OPZ3ZJBH/bundle.json","state":"https://pith.science/pith/K5H7VTRJZIBKMHCPQ3OPZ3ZJBH/state.json","well_known_bundle":"https://pith.science/.well-known/pith/K5H7VTRJZIBKMHCPQ3OPZ3ZJBH/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2024:K5H7VTRJZIBKMHCPQ3OPZ3ZJBH","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":"449c4945f771fd25b4e87a4308208b5add04c4782e3127c6ac0e4db85317bc21","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2024-10-02T18:20:51Z","title_canon_sha256":"806f15cab1868d82365f85ee75f05f34a95e7db58a442b764d52b3e3482dfb30"},"schema_version":"1.0","source":{"id":"2410.01926","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2410.01926","created_at":"2026-07-05T09:14:57Z"},{"alias_kind":"arxiv_version","alias_value":"2410.01926v1","created_at":"2026-07-05T09:14:57Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2410.01926","created_at":"2026-07-05T09:14:57Z"},{"alias_kind":"pith_short_12","alias_value":"K5H7VTRJZIBK","created_at":"2026-07-05T09:14:57Z"},{"alias_kind":"pith_short_16","alias_value":"K5H7VTRJZIBKMHCP","created_at":"2026-07-05T09:14:57Z"},{"alias_kind":"pith_short_8","alias_value":"K5H7VTRJ","created_at":"2026-07-05T09:14:57Z"}],"graph_snapshots":[{"event_id":"sha256:51f585f269000f9ef7e15a5840f6b7b5557a485f2d11f972d0be3713b9c5cf14","target":"graph","created_at":"2026-07-05T09:14:57Z","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/2410.01926/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Reconstructing past events requires reasoning across long time horizons. To figure out what happened, we need to use our prior knowledge about the world and human behavior and draw inferences from various sources of evidence including visual, language, and auditory cues. We introduce MARPLE, a benchmark for evaluating long-horizon inference capabilities using multi-modal evidence. Our benchmark features agents interacting with simulated households, supporting vision, language, and auditory stimuli, as well as procedurally generated environments and agent behaviors. Inspired by classic ``whodun","authors_text":"Emily Jin, Erik Brockbank, Hannah Cha, Jan-Philipp Fr\\\"anken, Jiajun Wu, Ruohan Zhang, Sarah Wu, Tobias Gerstenberg, Weiyu Liu, Zhuoyi Huang","cross_cats":[],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2024-10-02T18:20:51Z","title":"MARPLE: A Benchmark for Long-Horizon Inference"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2410.01926","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":"sha256:5a814b45c8cca326578c3b09bd5b23f9f966c2e38e3ac157d273af4321cf63ae","target":"record","created_at":"2026-07-05T09:14:57Z","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":"449c4945f771fd25b4e87a4308208b5add04c4782e3127c6ac0e4db85317bc21","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2024-10-02T18:20:51Z","title_canon_sha256":"806f15cab1868d82365f85ee75f05f34a95e7db58a442b764d52b3e3482dfb30"},"schema_version":"1.0","source":{"id":"2410.01926","kind":"arxiv","version":1}},"canonical_sha256":"574fface29ca02a61c4f86dcfcef2909e785000da4b9f7dbc792d248dfe25881","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"574fface29ca02a61c4f86dcfcef2909e785000da4b9f7dbc792d248dfe25881","first_computed_at":"2026-07-05T09:14:57.780408Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T09:14:57.780408Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"zVvCGS13Z0PqnN1/51exvZhQotn8WXQe4GOuq2xGeXLJmUugbUxj29e6DjJA5rVkR7HJFOKsUO+KhGWPhrgrAA==","signature_status":"signed_v1","signed_at":"2026-07-05T09:14:57.780820Z","signed_message":"canonical_sha256_bytes"},"source_id":"2410.01926","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:5a814b45c8cca326578c3b09bd5b23f9f966c2e38e3ac157d273af4321cf63ae","sha256:51f585f269000f9ef7e15a5840f6b7b5557a485f2d11f972d0be3713b9c5cf14"],"state_sha256":"b35903469d0d9f36777f95b800054bd347c6881ac1726c87345bfa8e17158e84"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"02MzDhrVY8LtYU8FYW37lJWw1NmIoXtXHOYT7I9paJ+Vjpwzz3rRLQ9yV/seySRM9CezFVlQ0Li6oUvcyRzSCw==","signed_message":"bundle_sha256_bytes","signed_at":"2026-07-07T07:10:13.620681Z","bundle_sha256":"0c7c7b8d066ffe0e504960fccab01c295772a78f3f9e9dca771b26aae2a34a21"}}