{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:52IGHM4VOCP357G3DHHHJNTUD5","short_pith_number":"pith:52IGHM4V","schema_version":"1.0","canonical_sha256":"ee9063b395709fbefcdb19ce74b6741f5efa8d1e0a944eff7bed2d991b74ae38","source":{"kind":"arxiv","id":"2605.13918","version":1},"attestation_state":"computed","paper":{"title":"CA2: Code-Aware Agent for Automated Game Testing","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"Reinforcement learning agents that observe call stack traces test games more effectively than agents limited to game state alone.","cross_cats":["cs.LG"],"primary_cat":"cs.SE","authors_text":"David Meger, Joshua Romoff, Valliappan Chidambaram Adaikkappan, Vincent Martineau","submitted_at":"2026-05-13T12:52:35Z","abstract_excerpt":"Automated game testing is important for verifying game functionality, but it remains a costly and time-consuming process. Manual testing often misses edge cases, and current automated methods struggle to provide full code coverage. Prior work has explored reinforcement learning (RL) for game testing, but without leveraging internal code signals such as the call stack. We present Code Aware Agent (CA2), which uses call stack information to learn effective testing strategies. The agent receives the current function call trace along with the game state and learns to reach specific target function"},"verification_status":{"content_addressed":true,"pith_receipt":true,"author_attested":false,"weak_author_claims":0,"strong_author_claims":0,"externally_anchored":false,"storage_verified":false,"citation_signatures":0,"replication_records":0,"graph_snapshot":true,"references_resolved":true,"formal_links_present":false},"canonical_record":{"source":{"id":"2605.13918","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.SE","submitted_at":"2026-05-13T12:52:35Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"c82ab882d0584c2c20c679ca7d122672c6f39060a2fb4e46a4b0b079142ba6c0","abstract_canon_sha256":"da6e7a52ec028f434af6eacf5bc284c1aeca146c4ef803fe4943992bcb0e6bbf"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-17T23:39:18.727155Z","signature_b64":"fmFwEepPj3d3C9FLWulaWqp8KIFs0yivZmT2QOv3Hn3TWhyinG+B9IfhLe/Ky3mj9p7xGfVx+t8vtD9l9LAZCw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"ee9063b395709fbefcdb19ce74b6741f5efa8d1e0a944eff7bed2d991b74ae38","last_reissued_at":"2026-05-17T23:39:18.726467Z","signature_status":"signed_v1","first_computed_at":"2026-05-17T23:39:18.726467Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"CA2: Code-Aware Agent for Automated Game Testing","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"Reinforcement learning agents that observe call stack traces test games more effectively than agents limited to game state alone.","cross_cats":["cs.LG"],"primary_cat":"cs.SE","authors_text":"David Meger, Joshua Romoff, Valliappan Chidambaram Adaikkappan, Vincent Martineau","submitted_at":"2026-05-13T12:52:35Z","abstract_excerpt":"Automated game testing is important for verifying game functionality, but it remains a costly and time-consuming process. Manual testing often misses edge cases, and current automated methods struggle to provide full code coverage. Prior work has explored reinforcement learning (RL) for game testing, but without leveraging internal code signals such as the call stack. We present Code Aware Agent (CA2), which uses call stack information to learn effective testing strategies. The agent receives the current function call trace along with the game state and learns to reach specific target function"},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"Our results show that incorporating code signals like the call stack enables more effective and targeted game testing.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"That the call stack can be extracted efficiently and that its addition to the observation space produces a genuine policy improvement rather than an artifact of the specific environments or reward design.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"CA2 integrates call stack information into RL agents for game testing and shows consistent gains over baselines that ignore code signals.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"Reinforcement learning agents that observe call stack traces test games more effectively than agents limited to game state alone.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"c59a26ef9101858973497123b842fdd709147c5d1a471a85372062bd04fc6cb1"},"source":{"id":"2605.13918","kind":"arxiv","version":1},"verdict":{"id":"e653964a-7b81-41a9-bfc9-995889f7a68c","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-15T05:49:50.226940Z","strongest_claim":"Our results show that incorporating code signals like the call stack enables more effective and targeted game testing.","one_line_summary":"CA2 integrates call stack information into RL agents for game testing and shows consistent gains over baselines that ignore code signals.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"That the call stack can be extracted efficiently and that its addition to the observation space produces a genuine policy improvement rather than an artifact of the specific environments or reward design.","pith_extraction_headline":"Reinforcement learning agents that observe call stack traces test games more effectively than agents limited to game state alone."},"references":{"count":33,"sample":[{"doi":"","year":2021,"title":"C. Politowski, F. Petrillo, and Y.-G. Guéhéneuc.A Survey of Video Game Testing. 2021. arXiv:2103.06431 [cs.SE].url:https://arxiv.org/abs/2103.06431","work_id":"17cd638a-68fb-40df-b62c-b10fc1b41d5e","ref_index":1,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"10.1109/cog57401.2023.10333194","year":2023,"title":"Technical Challenges of Deploying Reinforcement Learning Agents for Game Testing in AAA Games","work_id":"a8a2d7d2-80c8-410e-9264-02bd8ec3d61e","ref_index":2,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2013,"title":"Playing Atari with Deep Reinforcement Learning","work_id":"736a8ddf-e365-4940-ad58-4699fddedb86","ref_index":3,"cited_arxiv_id":"1312.5602","is_internal_anchor":true},{"doi":"","year":2024,"title":"Mastering Diverse Domains through World Models","work_id":"6aeb260f-8c7c-4f9c-b98b-067cd7c59acd","ref_index":4,"cited_arxiv_id":"2301.04104","is_internal_anchor":true},{"doi":"","year":2021,"title":"C. Gordillo, J. Bergdahl, K. Tollmar, and L. Gisslén.Improving Playtesting Coverage via Curiosity Driven Reinforcement Learning Agents. 2021. arXiv:2103 . 13798 [cs.LG].url: https://arxiv.org/abs/2103","work_id":"127e40d8-f3d5-417a-844d-8d63e4f1471a","ref_index":5,"cited_arxiv_id":"","is_internal_anchor":false}],"resolved_work":33,"snapshot_sha256":"110f0ea419bcc2636c0928a4941cd41fd0781aaf9b950cf6b381a5af1d42e5f6","internal_anchors":9},"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"},"aliases":[{"alias_kind":"arxiv","alias_value":"2605.13918","created_at":"2026-05-17T23:39:18.726591+00:00"},{"alias_kind":"arxiv_version","alias_value":"2605.13918v1","created_at":"2026-05-17T23:39:18.726591+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.13918","created_at":"2026-05-17T23:39:18.726591+00:00"},{"alias_kind":"pith_short_12","alias_value":"52IGHM4VOCP3","created_at":"2026-05-18T12:33:37.589309+00:00"},{"alias_kind":"pith_short_16","alias_value":"52IGHM4VOCP357G3","created_at":"2026-05-18T12:33:37.589309+00:00"},{"alias_kind":"pith_short_8","alias_value":"52IGHM4V","created_at":"2026-05-18T12:33:37.589309+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":0,"internal_anchor_count":0,"sample":[]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/52IGHM4VOCP357G3DHHHJNTUD5","json":"https://pith.science/pith/52IGHM4VOCP357G3DHHHJNTUD5.json","graph_json":"https://pith.science/api/pith-number/52IGHM4VOCP357G3DHHHJNTUD5/graph.json","events_json":"https://pith.science/api/pith-number/52IGHM4VOCP357G3DHHHJNTUD5/events.json","paper":"https://pith.science/paper/52IGHM4V"},"agent_actions":{"view_html":"https://pith.science/pith/52IGHM4VOCP357G3DHHHJNTUD5","download_json":"https://pith.science/pith/52IGHM4VOCP357G3DHHHJNTUD5.json","view_paper":"https://pith.science/paper/52IGHM4V","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2605.13918&json=true","fetch_graph":"https://pith.science/api/pith-number/52IGHM4VOCP357G3DHHHJNTUD5/graph.json","fetch_events":"https://pith.science/api/pith-number/52IGHM4VOCP357G3DHHHJNTUD5/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/52IGHM4VOCP357G3DHHHJNTUD5/action/timestamp_anchor","attest_storage":"https://pith.science/pith/52IGHM4VOCP357G3DHHHJNTUD5/action/storage_attestation","attest_author":"https://pith.science/pith/52IGHM4VOCP357G3DHHHJNTUD5/action/author_attestation","sign_citation":"https://pith.science/pith/52IGHM4VOCP357G3DHHHJNTUD5/action/citation_signature","submit_replication":"https://pith.science/pith/52IGHM4VOCP357G3DHHHJNTUD5/action/replication_record"}},"created_at":"2026-05-17T23:39:18.726591+00:00","updated_at":"2026-05-17T23:39:18.726591+00:00"}