{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2024:AFXWN6VWI2RWLLJR2IQI3VY2RJ","short_pith_number":"pith:AFXWN6VW","canonical_record":{"source":{"id":"2407.13943","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","primary_cat":"cs.CL","submitted_at":"2024-07-18T23:41:05Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"741c17bba9fd09235a558e89e0c7b20565dab764d422a7860f897bad07a1a318","abstract_canon_sha256":"b6a93bf8ec7f0038803d1656a0df955705c573079237303f3aa575138b39f710"},"schema_version":"1.0"},"canonical_sha256":"016f66fab646a365ad31d2208dd71a8a4c1d5c8318b8b80b7a28c27bc6560f80","source":{"kind":"arxiv","id":"2407.13943","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2407.13943","created_at":"2026-07-05T08:45:56Z"},{"alias_kind":"arxiv_version","alias_value":"2407.13943v1","created_at":"2026-07-05T08:45:56Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2407.13943","created_at":"2026-07-05T08:45:56Z"},{"alias_kind":"pith_short_12","alias_value":"AFXWN6VWI2RW","created_at":"2026-07-05T08:45:56Z"},{"alias_kind":"pith_short_16","alias_value":"AFXWN6VWI2RWLLJR","created_at":"2026-07-05T08:45:56Z"},{"alias_kind":"pith_short_8","alias_value":"AFXWN6VW","created_at":"2026-07-05T08:45:56Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2024:AFXWN6VWI2RWLLJR2IQI3VY2RJ","target":"record","payload":{"canonical_record":{"source":{"id":"2407.13943","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","primary_cat":"cs.CL","submitted_at":"2024-07-18T23:41:05Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"741c17bba9fd09235a558e89e0c7b20565dab764d422a7860f897bad07a1a318","abstract_canon_sha256":"b6a93bf8ec7f0038803d1656a0df955705c573079237303f3aa575138b39f710"},"schema_version":"1.0"},"canonical_sha256":"016f66fab646a365ad31d2208dd71a8a4c1d5c8318b8b80b7a28c27bc6560f80","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T08:45:56.011309Z","signature_b64":"Nwg5zVn1+5Aemsh74eTRufZ/oxyYPSd3OAi8Hi4EYekZZLMFTl3lkTB/IAzcMKQ5QobYKEwjOvXCYkIrMHAnBQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"016f66fab646a365ad31d2208dd71a8a4c1d5c8318b8b80b7a28c27bc6560f80","last_reissued_at":"2026-07-05T08:45:56.010770Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T08:45:56.010770Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2407.13943","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":"2026-07-05T08:45:56Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"tsCIrgIRsKmvFDV7bxx+ayR6byVVzXL4aJgHGaaWI/k12aCSp0HHYYpFwTFUgm/dxr0P7CpXNAEjHWj+E/0JCQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-06T06:13:06.122301Z"},"content_sha256":"0be818e9c8ad966a6cd8d2ffb5c0c000a2d188ca036eecf1bdd078a65bd14249","schema_version":"1.0","event_id":"sha256:0be818e9c8ad966a6cd8d2ffb5c0c000a2d188ca036eecf1bdd078a65bd14249"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2024:AFXWN6VWI2RWLLJR2IQI3VY2RJ","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Werewolf Arena: A Case Study in LLM Evaluation via Social Deduction","license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Feiyang Chen, Jane Friedhoff, Suma Bailis","submitted_at":"2024-07-18T23:41:05Z","abstract_excerpt":"This paper introduces Werewolf Arena, a novel framework for evaluating large language models (LLMs) through the lens of the classic social deduction game, Werewolf. In Werewolf Arena, LLMs compete against each other, navigating the game's complex dynamics of deception, deduction, and persuasion. The framework introduces a dynamic turn-taking system based on bidding, mirroring real-world discussions where individuals strategically choose when to speak. We demonstrate the framework's utility through an arena-style tournament featuring Gemini and GPT models. Our results reveal distinct strengths "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2407.13943","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/2407.13943/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":"2026-07-05T08:45:56Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"JMztjdB8SxU5IKbxmq0rh8n2yNQ+RNKtweg2dyyQcKpH3MeuiSlkx1vAJXruKkUv5MCVT6TH0JiFq6B3HGJfCw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-06T06:13:06.122790Z"},"content_sha256":"a7138bb9e3fec7aa3cac4e2e3616b3858deb26e852d6be23a499b22ac509d863","schema_version":"1.0","event_id":"sha256:a7138bb9e3fec7aa3cac4e2e3616b3858deb26e852d6be23a499b22ac509d863"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/AFXWN6VWI2RWLLJR2IQI3VY2RJ/bundle.json","state_url":"https://pith.science/pith/AFXWN6VWI2RWLLJR2IQI3VY2RJ/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/AFXWN6VWI2RWLLJR2IQI3VY2RJ/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":"2026-08-06T06:13:06Z","links":{"resolver":"https://pith.science/pith/AFXWN6VWI2RWLLJR2IQI3VY2RJ","bundle":"https://pith.science/pith/AFXWN6VWI2RWLLJR2IQI3VY2RJ/bundle.json","state":"https://pith.science/pith/AFXWN6VWI2RWLLJR2IQI3VY2RJ/state.json","well_known_bundle":"https://pith.science/.well-known/pith/AFXWN6VWI2RWLLJR2IQI3VY2RJ/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2024:AFXWN6VWI2RWLLJR2IQI3VY2RJ","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":"b6a93bf8ec7f0038803d1656a0df955705c573079237303f3aa575138b39f710","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","primary_cat":"cs.CL","submitted_at":"2024-07-18T23:41:05Z","title_canon_sha256":"741c17bba9fd09235a558e89e0c7b20565dab764d422a7860f897bad07a1a318"},"schema_version":"1.0","source":{"id":"2407.13943","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2407.13943","created_at":"2026-07-05T08:45:56Z"},{"alias_kind":"arxiv_version","alias_value":"2407.13943v1","created_at":"2026-07-05T08:45:56Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2407.13943","created_at":"2026-07-05T08:45:56Z"},{"alias_kind":"pith_short_12","alias_value":"AFXWN6VWI2RW","created_at":"2026-07-05T08:45:56Z"},{"alias_kind":"pith_short_16","alias_value":"AFXWN6VWI2RWLLJR","created_at":"2026-07-05T08:45:56Z"},{"alias_kind":"pith_short_8","alias_value":"AFXWN6VW","created_at":"2026-07-05T08:45:56Z"}],"graph_snapshots":[{"event_id":"sha256:a7138bb9e3fec7aa3cac4e2e3616b3858deb26e852d6be23a499b22ac509d863","target":"graph","created_at":"2026-07-05T08:45:56Z","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/2407.13943/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"This paper introduces Werewolf Arena, a novel framework for evaluating large language models (LLMs) through the lens of the classic social deduction game, Werewolf. In Werewolf Arena, LLMs compete against each other, navigating the game's complex dynamics of deception, deduction, and persuasion. The framework introduces a dynamic turn-taking system based on bidding, mirroring real-world discussions where individuals strategically choose when to speak. We demonstrate the framework's utility through an arena-style tournament featuring Gemini and GPT models. Our results reveal distinct strengths ","authors_text":"Feiyang Chen, Jane Friedhoff, Suma Bailis","cross_cats":["cs.AI"],"headline":"","license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","primary_cat":"cs.CL","submitted_at":"2024-07-18T23:41:05Z","title":"Werewolf Arena: A Case Study in LLM Evaluation via Social Deduction"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2407.13943","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":"sha256:0be818e9c8ad966a6cd8d2ffb5c0c000a2d188ca036eecf1bdd078a65bd14249","target":"record","created_at":"2026-07-05T08:45:56Z","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":"b6a93bf8ec7f0038803d1656a0df955705c573079237303f3aa575138b39f710","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","primary_cat":"cs.CL","submitted_at":"2024-07-18T23:41:05Z","title_canon_sha256":"741c17bba9fd09235a558e89e0c7b20565dab764d422a7860f897bad07a1a318"},"schema_version":"1.0","source":{"id":"2407.13943","kind":"arxiv","version":1}},"canonical_sha256":"016f66fab646a365ad31d2208dd71a8a4c1d5c8318b8b80b7a28c27bc6560f80","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"016f66fab646a365ad31d2208dd71a8a4c1d5c8318b8b80b7a28c27bc6560f80","first_computed_at":"2026-07-05T08:45:56.010770Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T08:45:56.010770Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"Nwg5zVn1+5Aemsh74eTRufZ/oxyYPSd3OAi8Hi4EYekZZLMFTl3lkTB/IAzcMKQ5QobYKEwjOvXCYkIrMHAnBQ==","signature_status":"signed_v1","signed_at":"2026-07-05T08:45:56.011309Z","signed_message":"canonical_sha256_bytes"},"source_id":"2407.13943","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:0be818e9c8ad966a6cd8d2ffb5c0c000a2d188ca036eecf1bdd078a65bd14249","sha256:a7138bb9e3fec7aa3cac4e2e3616b3858deb26e852d6be23a499b22ac509d863"],"state_sha256":"18e854a2a14fb7b3956986c88c9eef8006b3e6a46ba1fae1a9d1222ac71dd552"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"CobsXULMyNYpxHa1GCSl+eZFuYKfPaJ1agangliP+O+jsdL8iVaJbLpgn2X1ZGQv1YH2p4eC9r/GHMt1nR70Bw==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-06T06:13:06.126459Z","bundle_sha256":"8bb1cc880601058f85a0dddfd67943bf2634a62731907313de914b9708a2feb5"}}