{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2025:O22N6UK66VTDAST6PZ6XB5ISET","short_pith_number":"pith:O22N6UK6","canonical_record":{"source":{"id":"2508.08501","kind":"arxiv","version":3},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2025-08-11T22:17:07Z","cross_cats_sorted":[],"title_canon_sha256":"e7226e0c905d5cffb90fccff44cfe9aec2d570a99673a83ae88f51878e51e23f","abstract_canon_sha256":"882e1b9d5bc0dc0720756de6dd3a28255e62a927d757ac0ab027b4942a4a5fcc"},"schema_version":"1.0"},"canonical_sha256":"76b4df515ef566304a7e7e7d70f51224d83d1ddd2bd706817a6beb25ac6611e6","source":{"kind":"arxiv","id":"2508.08501","version":3},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2508.08501","created_at":"2026-05-20T00:02:05Z"},{"alias_kind":"arxiv_version","alias_value":"2508.08501v3","created_at":"2026-05-20T00:02:05Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2508.08501","created_at":"2026-05-20T00:02:05Z"},{"alias_kind":"pith_short_12","alias_value":"O22N6UK66VTD","created_at":"2026-05-20T00:02:05Z"},{"alias_kind":"pith_short_16","alias_value":"O22N6UK66VTDAST6","created_at":"2026-05-20T00:02:05Z"},{"alias_kind":"pith_short_8","alias_value":"O22N6UK6","created_at":"2026-05-20T00:02:05Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2025:O22N6UK66VTDAST6PZ6XB5ISET","target":"record","payload":{"canonical_record":{"source":{"id":"2508.08501","kind":"arxiv","version":3},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2025-08-11T22:17:07Z","cross_cats_sorted":[],"title_canon_sha256":"e7226e0c905d5cffb90fccff44cfe9aec2d570a99673a83ae88f51878e51e23f","abstract_canon_sha256":"882e1b9d5bc0dc0720756de6dd3a28255e62a927d757ac0ab027b4942a4a5fcc"},"schema_version":"1.0"},"canonical_sha256":"76b4df515ef566304a7e7e7d70f51224d83d1ddd2bd706817a6beb25ac6611e6","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-20T00:02:05.908921Z","signature_b64":"bsx+ZqK/zmpuKeAtrAqsk5LLlNCwAIAuUwJ/mgXmwFAobyt6Wo9Jty4Myfbo9RrkW+fCmq7qjTkyMbvhXAX1Bw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"76b4df515ef566304a7e7e7d70f51224d83d1ddd2bd706817a6beb25ac6611e6","last_reissued_at":"2026-05-20T00:02:05.908099Z","signature_status":"signed_v1","first_computed_at":"2026-05-20T00:02:05.908099Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2508.08501","source_version":3,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-05-20T00:02:05Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"RHoS1QLGTSfLMa5wTruaIa2wVrgzB9epEMtnZ4xD/vu2FwhQft2SEcyWLyB7l4Z58XuBNR+IrwJSumdLIenhCA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-07-01T07:46:00.239472Z"},"content_sha256":"1213eea2171e948bdeb172b220c243c1b589ac461c2fb1237ebfb2cb77540b50","schema_version":"1.0","event_id":"sha256:1213eea2171e948bdeb172b220c243c1b589ac461c2fb1237ebfb2cb77540b50"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2025:O22N6UK66VTDAST6PZ6XB5ISET","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"GVGAI-LLM: Evaluating Large Language Model Agents with Infinite Games","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.AI","authors_text":"Cong Lin, Jialin Liu, Julian Togelius, Muhammad Umair Nasir, Philip Bontrager, Yuchen Li","submitted_at":"2025-08-11T22:17:07Z","abstract_excerpt":"We introduce GVGAI-LLM, a video game benchmark for evaluating the reasoning and problem-solving capabilities of large language models (LLMs). Built on the General Video Game AI framework, it features a diverse collection of arcade-style games designed to test a model's ability to handle tasks that differ from most existing LLM benchmarks. The benchmark leverages a video game description language that enables the rapid creation of new games (including rules and levels), helping to prevent overfitting over time. Each game scene is represented by a compact set of ASCII characters, allowing for ef"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2508.08501","kind":"arxiv","version":3},"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/2508.08501/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":"2026-05-20T00:02:05Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"6sctHs5aKbtOj10RVs+1bBLWiENfjSEp1+TK9C8aq0pBeS8DqfKdzc/az/n09O/W+drbNo5CzU63E1SWN16HBA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-07-01T07:46:00.239892Z"},"content_sha256":"5824bd9f2e9cb58c21470c2c2310bce8d72a25dd1f6c9aab00fab352cd33c963","schema_version":"1.0","event_id":"sha256:5824bd9f2e9cb58c21470c2c2310bce8d72a25dd1f6c9aab00fab352cd33c963"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/O22N6UK66VTDAST6PZ6XB5ISET/bundle.json","state_url":"https://pith.science/pith/O22N6UK66VTDAST6PZ6XB5ISET/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/O22N6UK66VTDAST6PZ6XB5ISET/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":"2026-07-01T07:46:00Z","links":{"resolver":"https://pith.science/pith/O22N6UK66VTDAST6PZ6XB5ISET","bundle":"https://pith.science/pith/O22N6UK66VTDAST6PZ6XB5ISET/bundle.json","state":"https://pith.science/pith/O22N6UK66VTDAST6PZ6XB5ISET/state.json","well_known_bundle":"https://pith.science/.well-known/pith/O22N6UK66VTDAST6PZ6XB5ISET/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2025:O22N6UK66VTDAST6PZ6XB5ISET","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":"882e1b9d5bc0dc0720756de6dd3a28255e62a927d757ac0ab027b4942a4a5fcc","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2025-08-11T22:17:07Z","title_canon_sha256":"e7226e0c905d5cffb90fccff44cfe9aec2d570a99673a83ae88f51878e51e23f"},"schema_version":"1.0","source":{"id":"2508.08501","kind":"arxiv","version":3}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2508.08501","created_at":"2026-05-20T00:02:05Z"},{"alias_kind":"arxiv_version","alias_value":"2508.08501v3","created_at":"2026-05-20T00:02:05Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2508.08501","created_at":"2026-05-20T00:02:05Z"},{"alias_kind":"pith_short_12","alias_value":"O22N6UK66VTD","created_at":"2026-05-20T00:02:05Z"},{"alias_kind":"pith_short_16","alias_value":"O22N6UK66VTDAST6","created_at":"2026-05-20T00:02:05Z"},{"alias_kind":"pith_short_8","alias_value":"O22N6UK6","created_at":"2026-05-20T00:02:05Z"}],"graph_snapshots":[{"event_id":"sha256:5824bd9f2e9cb58c21470c2c2310bce8d72a25dd1f6c9aab00fab352cd33c963","target":"graph","created_at":"2026-05-20T00:02:05Z","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/2508.08501/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"We introduce GVGAI-LLM, a video game benchmark for evaluating the reasoning and problem-solving capabilities of large language models (LLMs). Built on the General Video Game AI framework, it features a diverse collection of arcade-style games designed to test a model's ability to handle tasks that differ from most existing LLM benchmarks. The benchmark leverages a video game description language that enables the rapid creation of new games (including rules and levels), helping to prevent overfitting over time. Each game scene is represented by a compact set of ASCII characters, allowing for ef","authors_text":"Cong Lin, Jialin Liu, Julian Togelius, Muhammad Umair Nasir, Philip Bontrager, Yuchen Li","cross_cats":[],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2025-08-11T22:17:07Z","title":"GVGAI-LLM: Evaluating Large Language Model Agents with Infinite Games"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2508.08501","kind":"arxiv","version":3},"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":"sha256:1213eea2171e948bdeb172b220c243c1b589ac461c2fb1237ebfb2cb77540b50","target":"record","created_at":"2026-05-20T00:02:05Z","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":"882e1b9d5bc0dc0720756de6dd3a28255e62a927d757ac0ab027b4942a4a5fcc","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2025-08-11T22:17:07Z","title_canon_sha256":"e7226e0c905d5cffb90fccff44cfe9aec2d570a99673a83ae88f51878e51e23f"},"schema_version":"1.0","source":{"id":"2508.08501","kind":"arxiv","version":3}},"canonical_sha256":"76b4df515ef566304a7e7e7d70f51224d83d1ddd2bd706817a6beb25ac6611e6","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"76b4df515ef566304a7e7e7d70f51224d83d1ddd2bd706817a6beb25ac6611e6","first_computed_at":"2026-05-20T00:02:05.908099Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-05-20T00:02:05.908099Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"bsx+ZqK/zmpuKeAtrAqsk5LLlNCwAIAuUwJ/mgXmwFAobyt6Wo9Jty4Myfbo9RrkW+fCmq7qjTkyMbvhXAX1Bw==","signature_status":"signed_v1","signed_at":"2026-05-20T00:02:05.908921Z","signed_message":"canonical_sha256_bytes"},"source_id":"2508.08501","source_kind":"arxiv","source_version":3}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:1213eea2171e948bdeb172b220c243c1b589ac461c2fb1237ebfb2cb77540b50","sha256:5824bd9f2e9cb58c21470c2c2310bce8d72a25dd1f6c9aab00fab352cd33c963"],"state_sha256":"0c7ddbe35aba95a1445e064aa0c77f8cb1da52776de2bf3093f44185d8d53f34"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"c3gbw+Sw7G/3jyUsPEBTZZRUOpwGMA67f1klbaww4k4NDiAs63/ka7ZwQ/BmO7Jx6x1/rWms6rCfMQqrWFYDDQ==","signed_message":"bundle_sha256_bytes","signed_at":"2026-07-01T07:46:00.241906Z","bundle_sha256":"040e901924fbab0e5a20b13b39fe709c87c762f0cfbf1e042946139c3585285b"}}