{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2026:C4X2WYUSJSGMQZH4MRZEAP2T3X","short_pith_number":"pith:C4X2WYUS","canonical_record":{"source":{"id":"2605.09098","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2026-05-09T18:12:10Z","cross_cats_sorted":[],"title_canon_sha256":"ab6532cfe56b9d38f446cf8e4a3280c49a48ae8c8aba8e6f1d481c1da2141a04","abstract_canon_sha256":"300f04c3e44cd403c909b99ad3164363d752ab3f313bfa35da1644359fd3544f"},"schema_version":"1.0"},"canonical_sha256":"172fab62924c8cc864fc6472403f53ddfb49c282c849cef40599428866edb749","source":{"kind":"arxiv","id":"2605.09098","version":2},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2605.09098","created_at":"2026-06-02T02:04:18Z"},{"alias_kind":"arxiv_version","alias_value":"2605.09098v2","created_at":"2026-06-02T02:04:18Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.09098","created_at":"2026-06-02T02:04:18Z"},{"alias_kind":"pith_short_12","alias_value":"C4X2WYUSJSGM","created_at":"2026-06-02T02:04:18Z"},{"alias_kind":"pith_short_16","alias_value":"C4X2WYUSJSGMQZH4","created_at":"2026-06-02T02:04:18Z"},{"alias_kind":"pith_short_8","alias_value":"C4X2WYUS","created_at":"2026-06-02T02:04:18Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2026:C4X2WYUSJSGMQZH4MRZEAP2T3X","target":"record","payload":{"canonical_record":{"source":{"id":"2605.09098","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2026-05-09T18:12:10Z","cross_cats_sorted":[],"title_canon_sha256":"ab6532cfe56b9d38f446cf8e4a3280c49a48ae8c8aba8e6f1d481c1da2141a04","abstract_canon_sha256":"300f04c3e44cd403c909b99ad3164363d752ab3f313bfa35da1644359fd3544f"},"schema_version":"1.0"},"canonical_sha256":"172fab62924c8cc864fc6472403f53ddfb49c282c849cef40599428866edb749","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-02T02:04:18.741492Z","signature_b64":"sqSwFeQoI860I1YpoweBbG9UM+CRY3Yxd3Oa7d1WdXkTS8pM929F0i4h6LpQ6mt4s+emn6Q+HesfZDE+5+zTAw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"172fab62924c8cc864fc6472403f53ddfb49c282c849cef40599428866edb749","last_reissued_at":"2026-06-02T02:04:18.741024Z","signature_status":"signed_v1","first_computed_at":"2026-06-02T02:04:18.741024Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2605.09098","source_version":2,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-06-02T02:04:18Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"FQjEgGec+1nzqDObVQmgOSnV/3cwVKws87CVXA/APKRU+ghr55xpPuj2MJ/NkgGAm/THvDSKZV1KzzOib2kXCA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-06T15:05:13.378256Z"},"content_sha256":"9db85311f72061a9fcd70298170b0037b8b811060b8e63991ebe7349f835adfc","schema_version":"1.0","event_id":"sha256:9db85311f72061a9fcd70298170b0037b8b811060b8e63991ebe7349f835adfc"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2026:C4X2WYUSJSGMQZH4MRZEAP2T3X","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Dynamic Meta-Metrics: Source-Sentence Conditioned Weighting for MT Evaluation","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"Source-sentence properties can guide adaptive combinations of machine translation metrics to better match human judgments.","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Aditya Khan, En-Shiun Annie Lee, Justin Vasselli, Luke Zhang, York Hay Ng","submitted_at":"2026-05-09T18:12:10Z","abstract_excerpt":"We propose Dynamic Meta-Metrics (DMM), a framework for machine translation evaluation that learns source-sentence conditioned combinations of existing metrics. Rather than relying on a single static ensemble or language-specific weighting, DMM adapts the metric combination based on properties of the source segment. We study hard conditioning, which fits an interpretable combiner per cluster, and an exploratory soft-conditioned extension whose weights vary continuously with source-cluster responsibilities. We evaluate DMM on the WMT Metrics Shared Task data across multiple language pairs using "},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"Across settings, MLP-based combinations outperform linear and Gaussian process-based ensembles, and introducing soft conditioning yields gains over linear models.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"That clustering or responsibility-based conditioning on source-sentence properties captures variations that meaningfully improve metric agreement with humans and generalizes to new data.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"Dynamic Meta-Metrics learns source-sentence-conditioned combinations of MT metrics, with MLP-based hard and soft clustering versions outperforming static linear and Gaussian process ensembles on WMT data.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"Source-sentence properties can guide adaptive combinations of machine translation metrics to better match human judgments.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"fd0fe60b642ccf147481f281e75eaa5a3d44c86bdc988d4f89fd5c23baa4c69c"},"source":{"id":"2605.09098","kind":"arxiv","version":2},"verdict":{"id":"e9fb3bf2-84bb-4f0f-8229-e55596a8691f","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-12T03:39:51.393536Z","strongest_claim":"Across settings, MLP-based combinations outperform linear and Gaussian process-based ensembles, and introducing soft conditioning yields gains over linear models.","one_line_summary":"Dynamic Meta-Metrics learns source-sentence-conditioned combinations of MT metrics, with MLP-based hard and soft clustering versions outperforming static linear and Gaussian process ensembles on WMT data.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"That clustering or responsibility-based conditioning on source-sentence properties captures variations that meaningfully improve metric agreement with humans and generalizes to new data.","pith_extraction_headline":"Source-sentence properties can guide adaptive combinations of machine translation metrics to better match human judgments."},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2605.09098/integrity.json","findings":[],"available":true,"detectors_run":[{"name":"claim_evidence","ran_at":"2026-05-20T08:22:01.672803Z","status":"completed","version":"1.0.0","findings_count":0},{"name":"ai_meta_artifact","ran_at":"2026-05-19T20:37:54.423626Z","status":"completed","version":"1.0.0","findings_count":0},{"name":"doi_title_agreement","ran_at":"2026-05-19T13:31:18.911066Z","status":"completed","version":"1.0.0","findings_count":0},{"name":"doi_compliance","ran_at":"2026-05-19T10:31:19.795995Z","status":"completed","version":"1.0.0","findings_count":0}],"snapshot_sha256":"b4bde15a0dca1552962f43a9670581a4ecd06883d24e88674aa1ebb66db2db55"},"references":{"count":0,"sample":[],"resolved_work":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","internal_anchors":0},"formal_canon":{"evidence_count":2,"snapshot_sha256":"bf6ee653a367a5fe138318d971efa492fb00ba51b12cab95478abf042d3ff1fc"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"},"verdict_id":"e9fb3bf2-84bb-4f0f-8229-e55596a8691f"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-06-02T02:04:18Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"Q424l0XAGkkMeKm3HwKYo0YK5ITyIubf8gNbaH4+pd0fLmeskwP58H3Tu0HoAh1yqRFJdelFvlZEUVDzch+MDg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-06T15:05:13.378978Z"},"content_sha256":"c33df63234673409236c67561c0410940ee97dfe951e0c9a7a5b1c7f56030b18","schema_version":"1.0","event_id":"sha256:c33df63234673409236c67561c0410940ee97dfe951e0c9a7a5b1c7f56030b18"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/C4X2WYUSJSGMQZH4MRZEAP2T3X/bundle.json","state_url":"https://pith.science/pith/C4X2WYUSJSGMQZH4MRZEAP2T3X/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/C4X2WYUSJSGMQZH4MRZEAP2T3X/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":"2026-08-06T15:05:13Z","links":{"resolver":"https://pith.science/pith/C4X2WYUSJSGMQZH4MRZEAP2T3X","bundle":"https://pith.science/pith/C4X2WYUSJSGMQZH4MRZEAP2T3X/bundle.json","state":"https://pith.science/pith/C4X2WYUSJSGMQZH4MRZEAP2T3X/state.json","well_known_bundle":"https://pith.science/.well-known/pith/C4X2WYUSJSGMQZH4MRZEAP2T3X/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:C4X2WYUSJSGMQZH4MRZEAP2T3X","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":"300f04c3e44cd403c909b99ad3164363d752ab3f313bfa35da1644359fd3544f","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2026-05-09T18:12:10Z","title_canon_sha256":"ab6532cfe56b9d38f446cf8e4a3280c49a48ae8c8aba8e6f1d481c1da2141a04"},"schema_version":"1.0","source":{"id":"2605.09098","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2605.09098","created_at":"2026-06-02T02:04:18Z"},{"alias_kind":"arxiv_version","alias_value":"2605.09098v2","created_at":"2026-06-02T02:04:18Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.09098","created_at":"2026-06-02T02:04:18Z"},{"alias_kind":"pith_short_12","alias_value":"C4X2WYUSJSGM","created_at":"2026-06-02T02:04:18Z"},{"alias_kind":"pith_short_16","alias_value":"C4X2WYUSJSGMQZH4","created_at":"2026-06-02T02:04:18Z"},{"alias_kind":"pith_short_8","alias_value":"C4X2WYUS","created_at":"2026-06-02T02:04:18Z"}],"graph_snapshots":[{"event_id":"sha256:c33df63234673409236c67561c0410940ee97dfe951e0c9a7a5b1c7f56030b18","target":"graph","created_at":"2026-06-02T02:04:18Z","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":4,"items":[{"attestation":"unclaimed","claim_id":"C1","kind":"strongest_claim","source":"verdict.strongest_claim","status":"machine_extracted","text":"Across settings, MLP-based combinations outperform linear and Gaussian process-based ensembles, and introducing soft conditioning yields gains over linear models."},{"attestation":"unclaimed","claim_id":"C2","kind":"weakest_assumption","source":"verdict.weakest_assumption","status":"machine_extracted","text":"That clustering or responsibility-based conditioning on source-sentence properties captures variations that meaningfully improve metric agreement with humans and generalizes to new data."},{"attestation":"unclaimed","claim_id":"C3","kind":"one_line_summary","source":"verdict.one_line_summary","status":"machine_extracted","text":"Dynamic Meta-Metrics learns source-sentence-conditioned combinations of MT metrics, with MLP-based hard and soft clustering versions outperforming static linear and Gaussian process ensembles on WMT data."},{"attestation":"unclaimed","claim_id":"C4","kind":"headline","source":"verdict.pith_extraction.headline","status":"machine_extracted","text":"Source-sentence properties can guide adaptive combinations of machine translation metrics to better match human judgments."}],"snapshot_sha256":"fd0fe60b642ccf147481f281e75eaa5a3d44c86bdc988d4f89fd5c23baa4c69c"},"formal_canon":{"evidence_count":2,"snapshot_sha256":"bf6ee653a367a5fe138318d971efa492fb00ba51b12cab95478abf042d3ff1fc"},"integrity":{"available":true,"clean":true,"detectors_run":[{"findings_count":0,"name":"claim_evidence","ran_at":"2026-05-20T08:22:01.672803Z","status":"completed","version":"1.0.0"},{"findings_count":0,"name":"ai_meta_artifact","ran_at":"2026-05-19T20:37:54.423626Z","status":"completed","version":"1.0.0"},{"findings_count":0,"name":"doi_title_agreement","ran_at":"2026-05-19T13:31:18.911066Z","status":"completed","version":"1.0.0"},{"findings_count":0,"name":"doi_compliance","ran_at":"2026-05-19T10:31:19.795995Z","status":"completed","version":"1.0.0"}],"endpoint":"/pith/2605.09098/integrity.json","findings":[],"snapshot_sha256":"b4bde15a0dca1552962f43a9670581a4ecd06883d24e88674aa1ebb66db2db55","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"We propose Dynamic Meta-Metrics (DMM), a framework for machine translation evaluation that learns source-sentence conditioned combinations of existing metrics. Rather than relying on a single static ensemble or language-specific weighting, DMM adapts the metric combination based on properties of the source segment. We study hard conditioning, which fits an interpretable combiner per cluster, and an exploratory soft-conditioned extension whose weights vary continuously with source-cluster responsibilities. We evaluate DMM on the WMT Metrics Shared Task data across multiple language pairs using ","authors_text":"Aditya Khan, En-Shiun Annie Lee, Justin Vasselli, Luke Zhang, York Hay Ng","cross_cats":[],"headline":"Source-sentence properties can guide adaptive combinations of machine translation metrics to better match human judgments.","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2026-05-09T18:12:10Z","title":"Dynamic Meta-Metrics: Source-Sentence Conditioned Weighting for MT Evaluation"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2605.09098","kind":"arxiv","version":2},"verdict":{"created_at":"2026-05-12T03:39:51.393536Z","id":"e9fb3bf2-84bb-4f0f-8229-e55596a8691f","model_set":{"reader":"grok-4.3"},"one_line_summary":"Dynamic Meta-Metrics learns source-sentence-conditioned combinations of MT metrics, with MLP-based hard and soft clustering versions outperforming static linear and Gaussian process ensembles on WMT data.","pipeline_version":"pith-pipeline@v0.9.0","pith_extraction_headline":"Source-sentence properties can guide adaptive combinations of machine translation metrics to better match human judgments.","strongest_claim":"Across settings, MLP-based combinations outperform linear and Gaussian process-based ensembles, and introducing soft conditioning yields gains over linear models.","weakest_assumption":"That clustering or responsibility-based conditioning on source-sentence properties captures variations that meaningfully improve metric agreement with humans and generalizes to new data."}},"verdict_id":"e9fb3bf2-84bb-4f0f-8229-e55596a8691f"}}],"author_attestations":[],"timestamp_anchors":[],"storage_attestations":[],"citation_signatures":[],"replication_records":[],"corrections":[],"mirror_hints":[],"record_created":{"event_id":"sha256:9db85311f72061a9fcd70298170b0037b8b811060b8e63991ebe7349f835adfc","target":"record","created_at":"2026-06-02T02:04:18Z","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":"300f04c3e44cd403c909b99ad3164363d752ab3f313bfa35da1644359fd3544f","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2026-05-09T18:12:10Z","title_canon_sha256":"ab6532cfe56b9d38f446cf8e4a3280c49a48ae8c8aba8e6f1d481c1da2141a04"},"schema_version":"1.0","source":{"id":"2605.09098","kind":"arxiv","version":2}},"canonical_sha256":"172fab62924c8cc864fc6472403f53ddfb49c282c849cef40599428866edb749","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"172fab62924c8cc864fc6472403f53ddfb49c282c849cef40599428866edb749","first_computed_at":"2026-06-02T02:04:18.741024Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-06-02T02:04:18.741024Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"sqSwFeQoI860I1YpoweBbG9UM+CRY3Yxd3Oa7d1WdXkTS8pM929F0i4h6LpQ6mt4s+emn6Q+HesfZDE+5+zTAw==","signature_status":"signed_v1","signed_at":"2026-06-02T02:04:18.741492Z","signed_message":"canonical_sha256_bytes"},"source_id":"2605.09098","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:9db85311f72061a9fcd70298170b0037b8b811060b8e63991ebe7349f835adfc","sha256:c33df63234673409236c67561c0410940ee97dfe951e0c9a7a5b1c7f56030b18"],"state_sha256":"33f73f28c4da3df32f3d35f2ac190096c3a45a90ed5534188b6bcc57e8f0bbad"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"Z4uzOwpAILorwtzk7N/WuF/BOn21ZK9TZgQrFM7dKryimSWjESppdwAPSYxHE9eYzZGm5cwmW5LtTX7US6VCBg==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-06T15:05:13.383853Z","bundle_sha256":"561f9a7630ec34117692f6c65cf808ecba2e025d6f22e6f4cd1cc905bc6712e3"}}