{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:ZH6L2A2H2NA7MMDSUZIBPFK2ZU","short_pith_number":"pith:ZH6L2A2H","schema_version":"1.0","canonical_sha256":"c9fcbd0347d341f63072a65017955acd04800f2be6ae9e38f7b19d04486d771e","source":{"kind":"arxiv","id":"2502.02289","version":1},"attestation_state":"computed","paper":{"title":"Evalita-LLM: Benchmarking Large Language Models on Italian","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Bernardo Magnini, Marco Madeddu, Martin Cimmino, Michele Resta, Paolo Albano, Roberto Zanoli, Viviana Patti","submitted_at":"2025-02-04T12:58:19Z","abstract_excerpt":"We describe Evalita-LLM, a new benchmark designed to evaluate Large Language Models (LLMs) on Italian tasks. The distinguishing and innovative features of Evalita-LLM are the following: (i) all tasks are native Italian, avoiding issues of translating from Italian and potential cultural biases; (ii) in addition to well established multiple-choice tasks, the benchmark includes generative tasks, enabling more natural interaction with LLMs; (iii) all tasks are evaluated against multiple prompts, this way mitigating the model sensitivity to specific prompts and allowing a fairer and objective evalu"},"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":"2502.02289","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2025-02-04T12:58:19Z","cross_cats_sorted":[],"title_canon_sha256":"ff67f7f904cfa5400f80fe76ff63c9baa489ceced62e57300ab6f24b0e6d80c3","abstract_canon_sha256":"75974ff34e032013fe7487cfbaf4b5eee5dfdd5eaedd0cd0e65a9a87152b7c18"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:09:26.586564Z","signature_b64":"D2tN4lFZUU9FbhnRbBYKg3O6r1a0S3H/SuXpC29uU1g/j1zM7i4qiT9XUxJ+fs6oAkDhBpMGk7gmLmkmrrKGCg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"c9fcbd0347d341f63072a65017955acd04800f2be6ae9e38f7b19d04486d771e","last_reissued_at":"2026-07-05T10:09:26.586055Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:09:26.586055Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Evalita-LLM: Benchmarking Large Language Models on Italian","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Bernardo Magnini, Marco Madeddu, Martin Cimmino, Michele Resta, Paolo Albano, Roberto Zanoli, Viviana Patti","submitted_at":"2025-02-04T12:58:19Z","abstract_excerpt":"We describe Evalita-LLM, a new benchmark designed to evaluate Large Language Models (LLMs) on Italian tasks. The distinguishing and innovative features of Evalita-LLM are the following: (i) all tasks are native Italian, avoiding issues of translating from Italian and potential cultural biases; (ii) in addition to well established multiple-choice tasks, the benchmark includes generative tasks, enabling more natural interaction with LLMs; (iii) all tasks are evaluated against multiple prompts, this way mitigating the model sensitivity to specific prompts and allowing a fairer and objective evalu"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2502.02289","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/2502.02289/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":"2502.02289","created_at":"2026-07-05T10:09:26.586120+00:00"},{"alias_kind":"arxiv_version","alias_value":"2502.02289v1","created_at":"2026-07-05T10:09:26.586120+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2502.02289","created_at":"2026-07-05T10:09:26.586120+00:00"},{"alias_kind":"pith_short_12","alias_value":"ZH6L2A2H2NA7","created_at":"2026-07-05T10:09:26.586120+00:00"},{"alias_kind":"pith_short_16","alias_value":"ZH6L2A2H2NA7MMDS","created_at":"2026-07-05T10:09:26.586120+00:00"},{"alias_kind":"pith_short_8","alias_value":"ZH6L2A2H","created_at":"2026-07-05T10:09:26.586120+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":2,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2605.07731","citing_title":"Benchmarking EngGPT2-16B-A3B against Comparable Italian and International Open-source LLMs","ref_index":83,"is_internal_anchor":false},{"citing_arxiv_id":"2605.07731","citing_title":"Benchmarking EngGPT2-16B-A3B against Comparable Italian and International Open-source LLMs","ref_index":83,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/ZH6L2A2H2NA7MMDSUZIBPFK2ZU","json":"https://pith.science/pith/ZH6L2A2H2NA7MMDSUZIBPFK2ZU.json","graph_json":"https://pith.science/api/pith-number/ZH6L2A2H2NA7MMDSUZIBPFK2ZU/graph.json","events_json":"https://pith.science/api/pith-number/ZH6L2A2H2NA7MMDSUZIBPFK2ZU/events.json","paper":"https://pith.science/paper/ZH6L2A2H"},"agent_actions":{"view_html":"https://pith.science/pith/ZH6L2A2H2NA7MMDSUZIBPFK2ZU","download_json":"https://pith.science/pith/ZH6L2A2H2NA7MMDSUZIBPFK2ZU.json","view_paper":"https://pith.science/paper/ZH6L2A2H","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2502.02289&json=true","fetch_graph":"https://pith.science/api/pith-number/ZH6L2A2H2NA7MMDSUZIBPFK2ZU/graph.json","fetch_events":"https://pith.science/api/pith-number/ZH6L2A2H2NA7MMDSUZIBPFK2ZU/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/ZH6L2A2H2NA7MMDSUZIBPFK2ZU/action/timestamp_anchor","attest_storage":"https://pith.science/pith/ZH6L2A2H2NA7MMDSUZIBPFK2ZU/action/storage_attestation","attest_author":"https://pith.science/pith/ZH6L2A2H2NA7MMDSUZIBPFK2ZU/action/author_attestation","sign_citation":"https://pith.science/pith/ZH6L2A2H2NA7MMDSUZIBPFK2ZU/action/citation_signature","submit_replication":"https://pith.science/pith/ZH6L2A2H2NA7MMDSUZIBPFK2ZU/action/replication_record"}},"created_at":"2026-07-05T10:09:26.586120+00:00","updated_at":"2026-07-05T10:09:26.586120+00:00"}