{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:UQZDXALHYY6ZA25CTICEHWQKV7","short_pith_number":"pith:UQZDXALH","schema_version":"1.0","canonical_sha256":"a4323b8167c63d906ba29a0443da0aaffcb42a818cee8ea92183e3539429be32","source":{"kind":"arxiv","id":"2410.13716","version":2},"attestation_state":"computed","paper":{"title":"MIRAGE-Bench: Automatic Multilingual Benchmark Arena for Retrieval-Augmented Generation Systems","license":"http://creativecommons.org/licenses/by-sa/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Amin Ahmad, Ge Luo, Jimmy Lin, Nandan Thakur, Suleman Kazi","submitted_at":"2024-10-17T16:18:49Z","abstract_excerpt":"Traditional retrieval-augmented generation (RAG) benchmarks evaluate systems using heuristic-based metrics, but these require human preferences as the ground truth for reference. In contrast, arena-based benchmarks, where systems compete against each other, require an expensive large language model (LLM) as a judge for a reliable evaluation. We present a simple efficient technique to combine the best of both worlds. The idea is to train a surrogate judge using heuristic metrics as input, to output the LLM as a judge prediction. In our work, we develop MIRAGE-Bench, a synthetic arena-based RAG "},"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":"2410.13716","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by-sa/4.0/","primary_cat":"cs.CL","submitted_at":"2024-10-17T16:18:49Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"c73d7e36c77310ddab14c7630539736f826227241d35822db08e8ce351195b86","abstract_canon_sha256":"5935a61411c037544233ebc3427f46911244d29f2ee0456264c531272591dce5"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:41:16.186622Z","signature_b64":"CpDuK9fh9J9oGMf2D42GRYdVr764VaFNOw6+eppcS5lkbLkxgxrB9q6BgQNPiIKLVXShs+VcDw+B0m9DF6k+AQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"a4323b8167c63d906ba29a0443da0aaffcb42a818cee8ea92183e3539429be32","last_reissued_at":"2026-07-05T10:41:16.186103Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:41:16.186103Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"MIRAGE-Bench: Automatic Multilingual Benchmark Arena for Retrieval-Augmented Generation Systems","license":"http://creativecommons.org/licenses/by-sa/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Amin Ahmad, Ge Luo, Jimmy Lin, Nandan Thakur, Suleman Kazi","submitted_at":"2024-10-17T16:18:49Z","abstract_excerpt":"Traditional retrieval-augmented generation (RAG) benchmarks evaluate systems using heuristic-based metrics, but these require human preferences as the ground truth for reference. In contrast, arena-based benchmarks, where systems compete against each other, require an expensive large language model (LLM) as a judge for a reliable evaluation. We present a simple efficient technique to combine the best of both worlds. The idea is to train a surrogate judge using heuristic metrics as input, to output the LLM as a judge prediction. In our work, we develop MIRAGE-Bench, a synthetic arena-based RAG "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2410.13716","kind":"arxiv","version":2},"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/2410.13716/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":"2410.13716","created_at":"2026-07-05T10:41:16.186174+00:00"},{"alias_kind":"arxiv_version","alias_value":"2410.13716v2","created_at":"2026-07-05T10:41:16.186174+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2410.13716","created_at":"2026-07-05T10:41:16.186174+00:00"},{"alias_kind":"pith_short_12","alias_value":"UQZDXALHYY6Z","created_at":"2026-07-05T10:41:16.186174+00:00"},{"alias_kind":"pith_short_16","alias_value":"UQZDXALHYY6ZA25C","created_at":"2026-07-05T10:41:16.186174+00:00"},{"alias_kind":"pith_short_8","alias_value":"UQZDXALH","created_at":"2026-07-05T10:41:16.186174+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2505.14226","citing_title":"Phonetic Perturbations Reveal Tokenizer-Rooted Safety Gaps in LLMs","ref_index":22,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/UQZDXALHYY6ZA25CTICEHWQKV7","json":"https://pith.science/pith/UQZDXALHYY6ZA25CTICEHWQKV7.json","graph_json":"https://pith.science/api/pith-number/UQZDXALHYY6ZA25CTICEHWQKV7/graph.json","events_json":"https://pith.science/api/pith-number/UQZDXALHYY6ZA25CTICEHWQKV7/events.json","paper":"https://pith.science/paper/UQZDXALH"},"agent_actions":{"view_html":"https://pith.science/pith/UQZDXALHYY6ZA25CTICEHWQKV7","download_json":"https://pith.science/pith/UQZDXALHYY6ZA25CTICEHWQKV7.json","view_paper":"https://pith.science/paper/UQZDXALH","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2410.13716&json=true","fetch_graph":"https://pith.science/api/pith-number/UQZDXALHYY6ZA25CTICEHWQKV7/graph.json","fetch_events":"https://pith.science/api/pith-number/UQZDXALHYY6ZA25CTICEHWQKV7/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/UQZDXALHYY6ZA25CTICEHWQKV7/action/timestamp_anchor","attest_storage":"https://pith.science/pith/UQZDXALHYY6ZA25CTICEHWQKV7/action/storage_attestation","attest_author":"https://pith.science/pith/UQZDXALHYY6ZA25CTICEHWQKV7/action/author_attestation","sign_citation":"https://pith.science/pith/UQZDXALHYY6ZA25CTICEHWQKV7/action/citation_signature","submit_replication":"https://pith.science/pith/UQZDXALHYY6ZA25CTICEHWQKV7/action/replication_record"}},"created_at":"2026-07-05T10:41:16.186174+00:00","updated_at":"2026-07-05T10:41:16.186174+00:00"}