{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2024:YB4CJT5WONCSOOMWVFMZ4KUHG3","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":"1d6720b9b0110bd43664b5b282f1816fbb288e696778c1111adcbea96de00a65","cross_cats_sorted":["cs.AI","cs.CV"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-10-09T10:09:11Z","title_canon_sha256":"26dc22e13f936a5203227a17e9bc9b2f8f3cd25f3f127fa2c6fdc37fc7ecfa39"},"schema_version":"1.0","source":{"id":"2410.06733","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2410.06733","created_at":"2026-07-05T09:18:06Z"},{"alias_kind":"arxiv_version","alias_value":"2410.06733v1","created_at":"2026-07-05T09:18:06Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2410.06733","created_at":"2026-07-05T09:18:06Z"},{"alias_kind":"pith_short_12","alias_value":"YB4CJT5WONCS","created_at":"2026-07-05T09:18:06Z"},{"alias_kind":"pith_short_16","alias_value":"YB4CJT5WONCSOOMW","created_at":"2026-07-05T09:18:06Z"},{"alias_kind":"pith_short_8","alias_value":"YB4CJT5W","created_at":"2026-07-05T09:18:06Z"}],"graph_snapshots":[{"event_id":"sha256:14cc0f666d510ece3412512ff3f466e030106786ce0513688080c79e0a280e1a","target":"graph","created_at":"2026-07-05T09:18:06Z","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/2410.06733/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"While advancements in NLP have significantly improved the performance of Large Language Models (LLMs) on tasks requiring vertical thinking, their lateral thinking capabilities remain under-explored and challenging to measure due to the complexity of assessing creative thought processes and the scarcity of relevant data. To address these challenges, we introduce SPLAT, a benchmark leveraging Situation Puzzles to evaluate and elicit LAteral Thinking of LLMs. This benchmark, containing 975 graded situation puzzles across three difficulty levels, employs a new multi-turn player-judge framework ins","authors_text":"Bowen Zhang, Gang Wang, Qi Chen, Qi Wu","cross_cats":["cs.AI","cs.CV"],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-10-09T10:09:11Z","title":"Weak-eval-Strong: Evaluating and Eliciting Lateral Thinking of LLMs with Situation Puzzles"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2410.06733","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":"sha256:4785bb5cd3a40a1665c1d37cf2937d0bb766166c3a3d7367d6b00d7bb853509d","target":"record","created_at":"2026-07-05T09:18:06Z","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":"1d6720b9b0110bd43664b5b282f1816fbb288e696778c1111adcbea96de00a65","cross_cats_sorted":["cs.AI","cs.CV"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-10-09T10:09:11Z","title_canon_sha256":"26dc22e13f936a5203227a17e9bc9b2f8f3cd25f3f127fa2c6fdc37fc7ecfa39"},"schema_version":"1.0","source":{"id":"2410.06733","kind":"arxiv","version":1}},"canonical_sha256":"c07824cfb67345273996a9599e2a8736e1368f97fb359bf93505967c09cdc8ed","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"c07824cfb67345273996a9599e2a8736e1368f97fb359bf93505967c09cdc8ed","first_computed_at":"2026-07-05T09:18:06.120789Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T09:18:06.120789Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"VxvHbTxyCGInSbrzng5meF+dRmSJlGH5rT+A8bfl7P7//kYzYr04ohszHLhU/ddvAjrFlNU9ODUnE8S7szhrBw==","signature_status":"signed_v1","signed_at":"2026-07-05T09:18:06.121192Z","signed_message":"canonical_sha256_bytes"},"source_id":"2410.06733","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:4785bb5cd3a40a1665c1d37cf2937d0bb766166c3a3d7367d6b00d7bb853509d","sha256:14cc0f666d510ece3412512ff3f466e030106786ce0513688080c79e0a280e1a"],"state_sha256":"77c4778dcdef6751232e6048c76defeb528490ac8fc2e38b6401fa60734aacb8"}