{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2026:6PWTYSPONUJGBKJUFZTOIPUJCZ","short_pith_number":"pith:6PWTYSPO","canonical_record":{"source":{"id":"2602.11103","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2026-02-11T18:15:11Z","cross_cats_sorted":["cs.CL","cs.SE"],"title_canon_sha256":"2f0ac47fd5dc9c0578656f6d6d825011b79d1860b85c00b9b99e1e568fbc71fc","abstract_canon_sha256":"f7814ca615865d22f5fcccd69ad70dc69b7e66a21d44d70c688b42613cff8dae"},"schema_version":"1.0"},"canonical_sha256":"f3ed3c49ee6d1260a9342e66e43e89166a90f0823c5cb766ffaa26e71988b588","source":{"kind":"arxiv","id":"2602.11103","version":2},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2602.11103","created_at":"2026-07-02T00:18:27Z"},{"alias_kind":"arxiv_version","alias_value":"2602.11103v2","created_at":"2026-07-02T00:18:27Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2602.11103","created_at":"2026-07-02T00:18:27Z"},{"alias_kind":"pith_short_12","alias_value":"6PWTYSPONUJG","created_at":"2026-07-02T00:18:27Z"},{"alias_kind":"pith_short_16","alias_value":"6PWTYSPONUJGBKJU","created_at":"2026-07-02T00:18:27Z"},{"alias_kind":"pith_short_8","alias_value":"6PWTYSPO","created_at":"2026-07-02T00:18:27Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2026:6PWTYSPONUJGBKJUFZTOIPUJCZ","target":"record","payload":{"canonical_record":{"source":{"id":"2602.11103","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2026-02-11T18:15:11Z","cross_cats_sorted":["cs.CL","cs.SE"],"title_canon_sha256":"2f0ac47fd5dc9c0578656f6d6d825011b79d1860b85c00b9b99e1e568fbc71fc","abstract_canon_sha256":"f7814ca615865d22f5fcccd69ad70dc69b7e66a21d44d70c688b42613cff8dae"},"schema_version":"1.0"},"canonical_sha256":"f3ed3c49ee6d1260a9342e66e43e89166a90f0823c5cb766ffaa26e71988b588","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-02T00:18:27.177164Z","signature_b64":"eYcycX1jASCl8akju4AmmPh/uLUHG5S3SZKuRC+CvpUVrBzy49lEOoEn2FR32VIwSwyJ1tbshA4gns3u0vMpCA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"f3ed3c49ee6d1260a9342e66e43e89166a90f0823c5cb766ffaa26e71988b588","last_reissued_at":"2026-07-02T00:18:27.176224Z","signature_status":"signed_v1","first_computed_at":"2026-07-02T00:18:27.176224Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2602.11103","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":"2026-07-02T00:18:27Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"VHc+etTOhaDsXI7G+0m/j40aVBZOrDAJdDKxYY4XYk6NZwajOmSpOFLLDsXnX0eZcU5c5M50R4dDvtT+s/LbDw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-02T02:54:52.193979Z"},"content_sha256":"1f8df8d03d8ec89f26567f371b65cf639e57bfbc90075d8032ab95b392266e73","schema_version":"1.0","event_id":"sha256:1f8df8d03d8ec89f26567f371b65cf639e57bfbc90075d8032ab95b392266e73"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2026:6PWTYSPONUJGBKJUFZTOIPUJCZ","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"GameDevBench: Evaluating Agentic Capabilities Through Game Development","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.CL","cs.SE"],"primary_cat":"cs.AI","authors_text":"Alexander Wang, Ameet Talwalkar, Arnav Yayavaram, Chris Donahue, Qiuhong Anna Wei, Runkun Chen, Seth Karten, Siddharth Yayavaram, Valerie Chen, Wayne Chi, Yixiong Fang","submitted_at":"2026-02-11T18:15:11Z","abstract_excerpt":"Despite rapid progress on coding agents, progress on their multimodal counterparts has lagged behind. A key challenge is the scarcity of evaluation testbeds that combine the complexity of software development with the need for deep multimodal understanding. In game development, agents must navigate large, dense codebases while manipulating intrinsically multimodal assets such as shaders, sprites, and animations within a visual game scene. We present GameDevBench, the first benchmark for evaluating agents on game development tasks. GameDevBench consists of 333 tasks derived from web and video t"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2602.11103","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/2602.11103/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":"2026-07-02T00:18:27Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"g0pzKEW3KYHeJl+QnjEczLbjJvLm7OAsxxce2kTkUmesHyO3FRFESsqaa5ITovHl91mihDIA0pTWmE6jNFUjBQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-02T02:54:52.194507Z"},"content_sha256":"bbb1d69a682b491d6e46b8b4a1d1f6269df2b5d36604f2d990790fc9e217f87f","schema_version":"1.0","event_id":"sha256:bbb1d69a682b491d6e46b8b4a1d1f6269df2b5d36604f2d990790fc9e217f87f"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/6PWTYSPONUJGBKJUFZTOIPUJCZ/bundle.json","state_url":"https://pith.science/pith/6PWTYSPONUJGBKJUFZTOIPUJCZ/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/6PWTYSPONUJGBKJUFZTOIPUJCZ/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":"2026-08-02T02:54:52Z","links":{"resolver":"https://pith.science/pith/6PWTYSPONUJGBKJUFZTOIPUJCZ","bundle":"https://pith.science/pith/6PWTYSPONUJGBKJUFZTOIPUJCZ/bundle.json","state":"https://pith.science/pith/6PWTYSPONUJGBKJUFZTOIPUJCZ/state.json","well_known_bundle":"https://pith.science/.well-known/pith/6PWTYSPONUJGBKJUFZTOIPUJCZ/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:6PWTYSPONUJGBKJUFZTOIPUJCZ","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":"f7814ca615865d22f5fcccd69ad70dc69b7e66a21d44d70c688b42613cff8dae","cross_cats_sorted":["cs.CL","cs.SE"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2026-02-11T18:15:11Z","title_canon_sha256":"2f0ac47fd5dc9c0578656f6d6d825011b79d1860b85c00b9b99e1e568fbc71fc"},"schema_version":"1.0","source":{"id":"2602.11103","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2602.11103","created_at":"2026-07-02T00:18:27Z"},{"alias_kind":"arxiv_version","alias_value":"2602.11103v2","created_at":"2026-07-02T00:18:27Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2602.11103","created_at":"2026-07-02T00:18:27Z"},{"alias_kind":"pith_short_12","alias_value":"6PWTYSPONUJG","created_at":"2026-07-02T00:18:27Z"},{"alias_kind":"pith_short_16","alias_value":"6PWTYSPONUJGBKJU","created_at":"2026-07-02T00:18:27Z"},{"alias_kind":"pith_short_8","alias_value":"6PWTYSPO","created_at":"2026-07-02T00:18:27Z"}],"graph_snapshots":[{"event_id":"sha256:bbb1d69a682b491d6e46b8b4a1d1f6269df2b5d36604f2d990790fc9e217f87f","target":"graph","created_at":"2026-07-02T00:18:27Z","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/2602.11103/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Despite rapid progress on coding agents, progress on their multimodal counterparts has lagged behind. A key challenge is the scarcity of evaluation testbeds that combine the complexity of software development with the need for deep multimodal understanding. In game development, agents must navigate large, dense codebases while manipulating intrinsically multimodal assets such as shaders, sprites, and animations within a visual game scene. We present GameDevBench, the first benchmark for evaluating agents on game development tasks. GameDevBench consists of 333 tasks derived from web and video t","authors_text":"Alexander Wang, Ameet Talwalkar, Arnav Yayavaram, Chris Donahue, Qiuhong Anna Wei, Runkun Chen, Seth Karten, Siddharth Yayavaram, Valerie Chen, Wayne Chi, Yixiong Fang","cross_cats":["cs.CL","cs.SE"],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2026-02-11T18:15:11Z","title":"GameDevBench: Evaluating Agentic Capabilities Through Game Development"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2602.11103","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":"sha256:1f8df8d03d8ec89f26567f371b65cf639e57bfbc90075d8032ab95b392266e73","target":"record","created_at":"2026-07-02T00:18:27Z","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":"f7814ca615865d22f5fcccd69ad70dc69b7e66a21d44d70c688b42613cff8dae","cross_cats_sorted":["cs.CL","cs.SE"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2026-02-11T18:15:11Z","title_canon_sha256":"2f0ac47fd5dc9c0578656f6d6d825011b79d1860b85c00b9b99e1e568fbc71fc"},"schema_version":"1.0","source":{"id":"2602.11103","kind":"arxiv","version":2}},"canonical_sha256":"f3ed3c49ee6d1260a9342e66e43e89166a90f0823c5cb766ffaa26e71988b588","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"f3ed3c49ee6d1260a9342e66e43e89166a90f0823c5cb766ffaa26e71988b588","first_computed_at":"2026-07-02T00:18:27.176224Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-02T00:18:27.176224Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"eYcycX1jASCl8akju4AmmPh/uLUHG5S3SZKuRC+CvpUVrBzy49lEOoEn2FR32VIwSwyJ1tbshA4gns3u0vMpCA==","signature_status":"signed_v1","signed_at":"2026-07-02T00:18:27.177164Z","signed_message":"canonical_sha256_bytes"},"source_id":"2602.11103","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:1f8df8d03d8ec89f26567f371b65cf639e57bfbc90075d8032ab95b392266e73","sha256:bbb1d69a682b491d6e46b8b4a1d1f6269df2b5d36604f2d990790fc9e217f87f"],"state_sha256":"83c942d52ab6cfbf09d23e80677d9360bde4ef96485e1bd0a46a6475762a30b3"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"IcS94xSbyr86YxB+SPybSwZvPCemWGN09E5dSr8460QgOJCG8E90qoIxUtsMy6kqtWavcNbW8DEPU0OTdBeEAA==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-02T02:54:52.200541Z","bundle_sha256":"e047765eea09cd4b9436680a7bc7eb35da098a4d1a5d0887d210fe198b3d2d06"}}