{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2024:DGCU4F4LOUXIU3IGAHMCJKM4SS","short_pith_number":"pith:DGCU4F4L","canonical_record":{"source":{"id":"2407.03129","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2024-07-03T14:12:04Z","cross_cats_sorted":[],"title_canon_sha256":"1c622ff1411f87e4c7c492ccc2061288b89a33903c3767cccba91415c10850c1","abstract_canon_sha256":"436f761cd9409ff25767d7ae20198190a8dee3f8111cfce0397349b6509d6558"},"schema_version":"1.0"},"canonical_sha256":"19854e178b752e8a6d0601d824a99c948faf2988e700e9022fbf4213e1913682","source":{"kind":"arxiv","id":"2407.03129","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2407.03129","created_at":"2026-07-05T08:39:46Z"},{"alias_kind":"arxiv_version","alias_value":"2407.03129v1","created_at":"2026-07-05T08:39:46Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2407.03129","created_at":"2026-07-05T08:39:46Z"},{"alias_kind":"pith_short_12","alias_value":"DGCU4F4LOUXI","created_at":"2026-07-05T08:39:46Z"},{"alias_kind":"pith_short_16","alias_value":"DGCU4F4LOUXIU3IG","created_at":"2026-07-05T08:39:46Z"},{"alias_kind":"pith_short_8","alias_value":"DGCU4F4L","created_at":"2026-07-05T08:39:46Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2024:DGCU4F4LOUXIU3IGAHMCJKM4SS","target":"record","payload":{"canonical_record":{"source":{"id":"2407.03129","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2024-07-03T14:12:04Z","cross_cats_sorted":[],"title_canon_sha256":"1c622ff1411f87e4c7c492ccc2061288b89a33903c3767cccba91415c10850c1","abstract_canon_sha256":"436f761cd9409ff25767d7ae20198190a8dee3f8111cfce0397349b6509d6558"},"schema_version":"1.0"},"canonical_sha256":"19854e178b752e8a6d0601d824a99c948faf2988e700e9022fbf4213e1913682","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T08:39:46.770235Z","signature_b64":"+ln4s9lvNJLeDxbuorT9TMChYsz0GJo5znEe26caHq8IQpUX1OOY8Qu6+obWa11zZOxuMiDcHHkDYWjrxD7iBA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"19854e178b752e8a6d0601d824a99c948faf2988e700e9022fbf4213e1913682","last_reissued_at":"2026-07-05T08:39:46.769772Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T08:39:46.769772Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2407.03129","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":"2026-07-05T08:39:46Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"zNuzYO16ksw37iX1LuBiGkFynEa4B4afNUy+u8miHoc2PboAkn+QieM1X7y4E1rLczReMSZ3i7VfdzwmlpNYBg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-07T19:49:11.441089Z"},"content_sha256":"2f381b39e660b19cbc83d860bec0388f019b3c1180e7f5cbe59505c62d10e869","schema_version":"1.0","event_id":"sha256:2f381b39e660b19cbc83d860bec0388f019b3c1180e7f5cbe59505c62d10e869"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2024:DGCU4F4LOUXIU3IGAHMCJKM4SS","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Social Bias Evaluation for Large Language Models Requires Prompt Variations","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Masahiro Kaneko, Naoaki Okazaki, Rem Hida","submitted_at":"2024-07-03T14:12:04Z","abstract_excerpt":"Warning: This paper contains examples of stereotypes and biases. Large Language Models (LLMs) exhibit considerable social biases, and various studies have tried to evaluate and mitigate these biases accurately. Previous studies use downstream tasks as prompts to examine the degree of social biases for evaluation and mitigation. While LLMs' output highly depends on prompts, previous studies evaluating and mitigating bias have often relied on a limited variety of prompts. In this paper, we investigate the sensitivity of LLMs when changing prompt variations (task instruction and prompt, few-shot "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2407.03129","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/2407.03129/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":"2026-07-05T08:39:46Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"kOAa55oQNti0jkgQc2jwCPVG5uOAjuhS4w0q5Ne4hwF2WEzFiAD+ujHv9myWPHz5q4RBqDodVgtc04qFAWHIAw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-07T19:49:11.441880Z"},"content_sha256":"5e7e779b82eeb05264faa3b760197b958f73090db5330b1f89f75431b71cfbd9","schema_version":"1.0","event_id":"sha256:5e7e779b82eeb05264faa3b760197b958f73090db5330b1f89f75431b71cfbd9"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/DGCU4F4LOUXIU3IGAHMCJKM4SS/bundle.json","state_url":"https://pith.science/pith/DGCU4F4LOUXIU3IGAHMCJKM4SS/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/DGCU4F4LOUXIU3IGAHMCJKM4SS/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":"2026-08-07T19:49:11Z","links":{"resolver":"https://pith.science/pith/DGCU4F4LOUXIU3IGAHMCJKM4SS","bundle":"https://pith.science/pith/DGCU4F4LOUXIU3IGAHMCJKM4SS/bundle.json","state":"https://pith.science/pith/DGCU4F4LOUXIU3IGAHMCJKM4SS/state.json","well_known_bundle":"https://pith.science/.well-known/pith/DGCU4F4LOUXIU3IGAHMCJKM4SS/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2024:DGCU4F4LOUXIU3IGAHMCJKM4SS","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":"436f761cd9409ff25767d7ae20198190a8dee3f8111cfce0397349b6509d6558","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2024-07-03T14:12:04Z","title_canon_sha256":"1c622ff1411f87e4c7c492ccc2061288b89a33903c3767cccba91415c10850c1"},"schema_version":"1.0","source":{"id":"2407.03129","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2407.03129","created_at":"2026-07-05T08:39:46Z"},{"alias_kind":"arxiv_version","alias_value":"2407.03129v1","created_at":"2026-07-05T08:39:46Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2407.03129","created_at":"2026-07-05T08:39:46Z"},{"alias_kind":"pith_short_12","alias_value":"DGCU4F4LOUXI","created_at":"2026-07-05T08:39:46Z"},{"alias_kind":"pith_short_16","alias_value":"DGCU4F4LOUXIU3IG","created_at":"2026-07-05T08:39:46Z"},{"alias_kind":"pith_short_8","alias_value":"DGCU4F4L","created_at":"2026-07-05T08:39:46Z"}],"graph_snapshots":[{"event_id":"sha256:5e7e779b82eeb05264faa3b760197b958f73090db5330b1f89f75431b71cfbd9","target":"graph","created_at":"2026-07-05T08:39:46Z","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/2407.03129/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Warning: This paper contains examples of stereotypes and biases. Large Language Models (LLMs) exhibit considerable social biases, and various studies have tried to evaluate and mitigate these biases accurately. Previous studies use downstream tasks as prompts to examine the degree of social biases for evaluation and mitigation. While LLMs' output highly depends on prompts, previous studies evaluating and mitigating bias have often relied on a limited variety of prompts. In this paper, we investigate the sensitivity of LLMs when changing prompt variations (task instruction and prompt, few-shot ","authors_text":"Masahiro Kaneko, Naoaki Okazaki, Rem Hida","cross_cats":[],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2024-07-03T14:12:04Z","title":"Social Bias Evaluation for Large Language Models Requires Prompt Variations"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2407.03129","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":"sha256:2f381b39e660b19cbc83d860bec0388f019b3c1180e7f5cbe59505c62d10e869","target":"record","created_at":"2026-07-05T08:39:46Z","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":"436f761cd9409ff25767d7ae20198190a8dee3f8111cfce0397349b6509d6558","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2024-07-03T14:12:04Z","title_canon_sha256":"1c622ff1411f87e4c7c492ccc2061288b89a33903c3767cccba91415c10850c1"},"schema_version":"1.0","source":{"id":"2407.03129","kind":"arxiv","version":1}},"canonical_sha256":"19854e178b752e8a6d0601d824a99c948faf2988e700e9022fbf4213e1913682","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"19854e178b752e8a6d0601d824a99c948faf2988e700e9022fbf4213e1913682","first_computed_at":"2026-07-05T08:39:46.769772Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T08:39:46.769772Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"+ln4s9lvNJLeDxbuorT9TMChYsz0GJo5znEe26caHq8IQpUX1OOY8Qu6+obWa11zZOxuMiDcHHkDYWjrxD7iBA==","signature_status":"signed_v1","signed_at":"2026-07-05T08:39:46.770235Z","signed_message":"canonical_sha256_bytes"},"source_id":"2407.03129","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:2f381b39e660b19cbc83d860bec0388f019b3c1180e7f5cbe59505c62d10e869","sha256:5e7e779b82eeb05264faa3b760197b958f73090db5330b1f89f75431b71cfbd9"],"state_sha256":"857fdb3ae1c8daab2ca2da6e11eb9aa0182b49615a923dba4fa7a9ee542f228f"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"24UmSQpCILVF5/s2Wv3lw2lbp+bzAl24K02qflA5OHQYbXvHZyYGNa83po0jl7WiLL3sliM+ehON4Yb0G7ZyDA==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-07T19:49:11.449220Z","bundle_sha256":"192e5f6ba3a1ae7c63384fadcaa83a1f8e2f375116c845d56fbba4458fe8b58d"}}