{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:O3EDJVYLBEMKKQVWEWFJNVDWWK","short_pith_number":"pith:O3EDJVYL","schema_version":"1.0","canonical_sha256":"76c834d70b0918a542b6258a96d476b2959d410a46e2c0265e359c6b238c7463","source":{"kind":"arxiv","id":"2505.19776","version":1},"attestation_state":"computed","paper":{"title":"Analyzing Political Bias in LLMs via Target-Oriented Sentiment Classification","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Adrian Popescu, Akram Elbouanani, Evan Dufraisse","submitted_at":"2025-05-26T10:01:24Z","abstract_excerpt":"Political biases encoded by LLMs might have detrimental effects on downstream applications. Existing bias analysis methods rely on small-size intermediate tasks (questionnaire answering or political content generation) and rely on the LLMs themselves for analysis, thus propagating bias. We propose a new approach leveraging the observation that LLM sentiment predictions vary with the target entity in the same sentence. We define an entropy-based inconsistency metric to encode this prediction variability. We insert 1319 demographically and politically diverse politician names in 450 political se"},"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":false,"formal_links_present":false},"canonical_record":{"source":{"id":"2505.19776","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-05-26T10:01:24Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"7462efeabd641691faa03b41bb1bcca6d025f8874f2b9ad50374f5a7f215f99f","abstract_canon_sha256":"8090f9c5eff479e411d4b776dd2a1864efc5cd37b439baa7a2ea1a034521608c"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:09:35.999031Z","signature_b64":"zk8idT686+cn82j0jXab8Kptys9KuwiWe0iSB4oVxPIAelNSg1i/fHVB2U39xzJmXEFNxgjb4Uv2d4sQOMlxCw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"76c834d70b0918a542b6258a96d476b2959d410a46e2c0265e359c6b238c7463","last_reissued_at":"2026-07-05T11:09:35.998543Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:09:35.998543Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Analyzing Political Bias in LLMs via Target-Oriented Sentiment Classification","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Adrian Popescu, Akram Elbouanani, Evan Dufraisse","submitted_at":"2025-05-26T10:01:24Z","abstract_excerpt":"Political biases encoded by LLMs might have detrimental effects on downstream applications. Existing bias analysis methods rely on small-size intermediate tasks (questionnaire answering or political content generation) and rely on the LLMs themselves for analysis, thus propagating bias. We propose a new approach leveraging the observation that LLM sentiment predictions vary with the target entity in the same sentence. We define an entropy-based inconsistency metric to encode this prediction variability. We insert 1319 demographically and politically diverse politician names in 450 political se"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2505.19776","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/2505.19776/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"},"aliases":[{"alias_kind":"arxiv","alias_value":"2505.19776","created_at":"2026-07-05T11:09:35.998603+00:00"},{"alias_kind":"arxiv_version","alias_value":"2505.19776v1","created_at":"2026-07-05T11:09:35.998603+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2505.19776","created_at":"2026-07-05T11:09:35.998603+00:00"},{"alias_kind":"pith_short_12","alias_value":"O3EDJVYLBEMK","created_at":"2026-07-05T11:09:35.998603+00:00"},{"alias_kind":"pith_short_16","alias_value":"O3EDJVYLBEMKKQVW","created_at":"2026-07-05T11:09:35.998603+00:00"},{"alias_kind":"pith_short_8","alias_value":"O3EDJVYL","created_at":"2026-07-05T11:09:35.998603+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/O3EDJVYLBEMKKQVWEWFJNVDWWK","json":"https://pith.science/pith/O3EDJVYLBEMKKQVWEWFJNVDWWK.json","graph_json":"https://pith.science/api/pith-number/O3EDJVYLBEMKKQVWEWFJNVDWWK/graph.json","events_json":"https://pith.science/api/pith-number/O3EDJVYLBEMKKQVWEWFJNVDWWK/events.json","paper":"https://pith.science/paper/O3EDJVYL"},"agent_actions":{"view_html":"https://pith.science/pith/O3EDJVYLBEMKKQVWEWFJNVDWWK","download_json":"https://pith.science/pith/O3EDJVYLBEMKKQVWEWFJNVDWWK.json","view_paper":"https://pith.science/paper/O3EDJVYL","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2505.19776&json=true","fetch_graph":"https://pith.science/api/pith-number/O3EDJVYLBEMKKQVWEWFJNVDWWK/graph.json","fetch_events":"https://pith.science/api/pith-number/O3EDJVYLBEMKKQVWEWFJNVDWWK/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/O3EDJVYLBEMKKQVWEWFJNVDWWK/action/timestamp_anchor","attest_storage":"https://pith.science/pith/O3EDJVYLBEMKKQVWEWFJNVDWWK/action/storage_attestation","attest_author":"https://pith.science/pith/O3EDJVYLBEMKKQVWEWFJNVDWWK/action/author_attestation","sign_citation":"https://pith.science/pith/O3EDJVYLBEMKKQVWEWFJNVDWWK/action/citation_signature","submit_replication":"https://pith.science/pith/O3EDJVYLBEMKKQVWEWFJNVDWWK/action/replication_record"}},"created_at":"2026-07-05T11:09:35.998603+00:00","updated_at":"2026-07-05T11:09:35.998603+00:00"}