{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2025:P6K4LUFN7XHUNAAJN7GBDMYAVV","short_pith_number":"pith:P6K4LUFN","canonical_record":{"source":{"id":"2502.18431","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-02-25T18:26:48Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"1cfb14394b08eab3c49680d61d1dda1f003d4bd39efd9aaaeb589e56a2829336","abstract_canon_sha256":"85e8e78ac16b000a66fbb0a6a2d42aeab9e6851cf7070f6fc47836b3f3b36773"},"schema_version":"1.0"},"canonical_sha256":"7f95c5d0adfdcf4680096fcc11b300ad7bf8a4c8baefc69d55f3a45acea7cac2","source":{"kind":"arxiv","id":"2502.18431","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2502.18431","created_at":"2026-07-05T10:19:53Z"},{"alias_kind":"arxiv_version","alias_value":"2502.18431v1","created_at":"2026-07-05T10:19:53Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2502.18431","created_at":"2026-07-05T10:19:53Z"},{"alias_kind":"pith_short_12","alias_value":"P6K4LUFN7XHU","created_at":"2026-07-05T10:19:53Z"},{"alias_kind":"pith_short_16","alias_value":"P6K4LUFN7XHUNAAJ","created_at":"2026-07-05T10:19:53Z"},{"alias_kind":"pith_short_8","alias_value":"P6K4LUFN","created_at":"2026-07-05T10:19:53Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2025:P6K4LUFN7XHUNAAJN7GBDMYAVV","target":"record","payload":{"canonical_record":{"source":{"id":"2502.18431","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-02-25T18:26:48Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"1cfb14394b08eab3c49680d61d1dda1f003d4bd39efd9aaaeb589e56a2829336","abstract_canon_sha256":"85e8e78ac16b000a66fbb0a6a2d42aeab9e6851cf7070f6fc47836b3f3b36773"},"schema_version":"1.0"},"canonical_sha256":"7f95c5d0adfdcf4680096fcc11b300ad7bf8a4c8baefc69d55f3a45acea7cac2","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:19:53.677513Z","signature_b64":"nVF9ipuG5kUBOzNFNVos2mt+Tg9WUDLl6kldrOpeV5PBjlI0b9EuqknFWwJhgEocXhOEpBN+//p7VaaDWTuPCQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"7f95c5d0adfdcf4680096fcc11b300ad7bf8a4c8baefc69d55f3a45acea7cac2","last_reissued_at":"2026-07-05T10:19:53.677026Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:19:53.677026Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2502.18431","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":"2026-07-05T10:19:53Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"lb2LdUYKZtzLxf243e15+qC0wKofUhlpjnDynMP0z+cIljSINdGnDng8Vs3jdjN+LIZXyWVEMF/a34r4zTQ/Aw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-15T07:14:47.723202Z"},"content_sha256":"7bdba844c4f782a9b9a855f905335b46964e6ed6f5f3a377127f21884fe4342b","schema_version":"1.0","event_id":"sha256:7bdba844c4f782a9b9a855f905335b46964e6ed6f5f3a377127f21884fe4342b"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2025:P6K4LUFN7XHUNAAJN7GBDMYAVV","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"TextGames: Learning to Self-Play Text-Based Puzzle Games via Language Model Reasoning","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Alham Fikri Aji, Frederikus Hudi, Genta Indra Winata, Ruochen Zhang","submitted_at":"2025-02-25T18:26:48Z","abstract_excerpt":"Reasoning is a fundamental capability of large language models (LLMs), enabling them to comprehend, analyze, and solve complex problems. In this paper, we introduce TextGames, an innovative benchmark specifically crafted to assess LLMs through demanding text-based games that require advanced skills in pattern recognition, spatial awareness, arithmetic, and logical reasoning. Our analysis probes LLMs' performance in both single-turn and multi-turn reasoning, and their abilities in leveraging feedback to correct subsequent answers through self-reflection. Our findings reveal that, although LLMs "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2502.18431","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/2502.18431/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":"2026-07-05T10:19:53Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"Y0YwA71CdiIzDoJ5WfbORJBV99UaAg3sqd0hE7/1Fpa3IVobiz+u44r2mXxALh6tb52KicTWDT1fd/qaAAyTCA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-15T07:14:47.723748Z"},"content_sha256":"431edf37cd264938fdba02f7cc6d634170143f5b37473053b477a57fa5abaa1e","schema_version":"1.0","event_id":"sha256:431edf37cd264938fdba02f7cc6d634170143f5b37473053b477a57fa5abaa1e"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/P6K4LUFN7XHUNAAJN7GBDMYAVV/bundle.json","state_url":"https://pith.science/pith/P6K4LUFN7XHUNAAJN7GBDMYAVV/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/P6K4LUFN7XHUNAAJN7GBDMYAVV/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":"2026-08-15T07:14:47Z","links":{"resolver":"https://pith.science/pith/P6K4LUFN7XHUNAAJN7GBDMYAVV","bundle":"https://pith.science/pith/P6K4LUFN7XHUNAAJN7GBDMYAVV/bundle.json","state":"https://pith.science/pith/P6K4LUFN7XHUNAAJN7GBDMYAVV/state.json","well_known_bundle":"https://pith.science/.well-known/pith/P6K4LUFN7XHUNAAJN7GBDMYAVV/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2025:P6K4LUFN7XHUNAAJN7GBDMYAVV","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":"85e8e78ac16b000a66fbb0a6a2d42aeab9e6851cf7070f6fc47836b3f3b36773","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-02-25T18:26:48Z","title_canon_sha256":"1cfb14394b08eab3c49680d61d1dda1f003d4bd39efd9aaaeb589e56a2829336"},"schema_version":"1.0","source":{"id":"2502.18431","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2502.18431","created_at":"2026-07-05T10:19:53Z"},{"alias_kind":"arxiv_version","alias_value":"2502.18431v1","created_at":"2026-07-05T10:19:53Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2502.18431","created_at":"2026-07-05T10:19:53Z"},{"alias_kind":"pith_short_12","alias_value":"P6K4LUFN7XHU","created_at":"2026-07-05T10:19:53Z"},{"alias_kind":"pith_short_16","alias_value":"P6K4LUFN7XHUNAAJ","created_at":"2026-07-05T10:19:53Z"},{"alias_kind":"pith_short_8","alias_value":"P6K4LUFN","created_at":"2026-07-05T10:19:53Z"}],"graph_snapshots":[{"event_id":"sha256:431edf37cd264938fdba02f7cc6d634170143f5b37473053b477a57fa5abaa1e","target":"graph","created_at":"2026-07-05T10:19:53Z","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/2502.18431/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Reasoning is a fundamental capability of large language models (LLMs), enabling them to comprehend, analyze, and solve complex problems. In this paper, we introduce TextGames, an innovative benchmark specifically crafted to assess LLMs through demanding text-based games that require advanced skills in pattern recognition, spatial awareness, arithmetic, and logical reasoning. Our analysis probes LLMs' performance in both single-turn and multi-turn reasoning, and their abilities in leveraging feedback to correct subsequent answers through self-reflection. Our findings reveal that, although LLMs ","authors_text":"Alham Fikri Aji, Frederikus Hudi, Genta Indra Winata, Ruochen Zhang","cross_cats":["cs.AI"],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-02-25T18:26:48Z","title":"TextGames: Learning to Self-Play Text-Based Puzzle Games via Language Model Reasoning"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2502.18431","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":"sha256:7bdba844c4f782a9b9a855f905335b46964e6ed6f5f3a377127f21884fe4342b","target":"record","created_at":"2026-07-05T10:19:53Z","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":"85e8e78ac16b000a66fbb0a6a2d42aeab9e6851cf7070f6fc47836b3f3b36773","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-02-25T18:26:48Z","title_canon_sha256":"1cfb14394b08eab3c49680d61d1dda1f003d4bd39efd9aaaeb589e56a2829336"},"schema_version":"1.0","source":{"id":"2502.18431","kind":"arxiv","version":1}},"canonical_sha256":"7f95c5d0adfdcf4680096fcc11b300ad7bf8a4c8baefc69d55f3a45acea7cac2","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"7f95c5d0adfdcf4680096fcc11b300ad7bf8a4c8baefc69d55f3a45acea7cac2","first_computed_at":"2026-07-05T10:19:53.677026Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T10:19:53.677026Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"nVF9ipuG5kUBOzNFNVos2mt+Tg9WUDLl6kldrOpeV5PBjlI0b9EuqknFWwJhgEocXhOEpBN+//p7VaaDWTuPCQ==","signature_status":"signed_v1","signed_at":"2026-07-05T10:19:53.677513Z","signed_message":"canonical_sha256_bytes"},"source_id":"2502.18431","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:7bdba844c4f782a9b9a855f905335b46964e6ed6f5f3a377127f21884fe4342b","sha256:431edf37cd264938fdba02f7cc6d634170143f5b37473053b477a57fa5abaa1e"],"state_sha256":"749ceb1ee32e53868a1acbe431fc832dcc101acfea97f5e43de29084b8899b19"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"PQevDJwbEhaLTuqqr1Ld6B6KLdleULAvt0YdIW8kK9lFfHFQny4nrqMxCzVZ7o+fziwwZbCb1mWwlUTf9mZmAg==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-15T07:14:47.729954Z","bundle_sha256":"b6c5df32c20cf60fe8c67517363b8649ea06af48f360db2ddb59e779693b476e"}}