{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:CZ6EL7JCVVXAZRVJ5KCH2KSXI7","short_pith_number":"pith:CZ6EL7JC","schema_version":"1.0","canonical_sha256":"167c45fd22ad6e0cc6a9ea847d2a5747d5db2252a3292120f78a001111eaaaee","source":{"kind":"arxiv","id":"2505.24857","version":1},"attestation_state":"computed","paper":{"title":"Accelerated Sampling from Masked Diffusion Models via Entropy Bounded Unmasking","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Brian Karrer, Daniel Severo, Heli Ben-Hamu, Itai Gat, Niklas Nolte","submitted_at":"2025-05-30T17:52:55Z","abstract_excerpt":"Recent masked diffusion models (MDMs) have shown competitive performance compared to autoregressive models (ARMs) for language modeling. While most literature has focused on performance enhancing sampling procedures, efficient sampling from MDMs has been scarcely explored. We make the observation that often a given sequence of partially masked tokens determines the values of multiple unknown tokens deterministically, meaning that a single prediction of a masked model holds additional information unused by standard sampling procedures. Based on this observation, we introduce EB-Sampler, a simpl"},"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":"2505.24857","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2025-05-30T17:52:55Z","cross_cats_sorted":[],"title_canon_sha256":"8617a5b6b1a1489d5654497e168c70c7a869d1bdc553f287cb69235b936498ba","abstract_canon_sha256":"1b8088ad5d5b094a132474033fa741d5dc4fbb4f4db7da96ab3da81529c727ae"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:12:57.898197Z","signature_b64":"dx5Fvl1QHrMitxfsfaxWGSku845oGgYaqCnrsQjewCCWcwO9S0byY0eBIjsSmAtl6C/l4DpI0NlACzNsuK/oAw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"167c45fd22ad6e0cc6a9ea847d2a5747d5db2252a3292120f78a001111eaaaee","last_reissued_at":"2026-07-05T11:12:57.897584Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:12:57.897584Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Accelerated Sampling from Masked Diffusion Models via Entropy Bounded Unmasking","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Brian Karrer, Daniel Severo, Heli Ben-Hamu, Itai Gat, Niklas Nolte","submitted_at":"2025-05-30T17:52:55Z","abstract_excerpt":"Recent masked diffusion models (MDMs) have shown competitive performance compared to autoregressive models (ARMs) for language modeling. While most literature has focused on performance enhancing sampling procedures, efficient sampling from MDMs has been scarcely explored. We make the observation that often a given sequence of partially masked tokens determines the values of multiple unknown tokens deterministically, meaning that a single prediction of a masked model holds additional information unused by standard sampling procedures. Based on this observation, we introduce EB-Sampler, a simpl"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2505.24857","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/2505.24857/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":"2505.24857","created_at":"2026-07-05T11:12:57.897652+00:00"},{"alias_kind":"arxiv_version","alias_value":"2505.24857v1","created_at":"2026-07-05T11:12:57.897652+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2505.24857","created_at":"2026-07-05T11:12:57.897652+00:00"},{"alias_kind":"pith_short_12","alias_value":"CZ6EL7JCVVXA","created_at":"2026-07-05T11:12:57.897652+00:00"},{"alias_kind":"pith_short_16","alias_value":"CZ6EL7JCVVXAZRVJ","created_at":"2026-07-05T11:12:57.897652+00:00"},{"alias_kind":"pith_short_8","alias_value":"CZ6EL7JC","created_at":"2026-07-05T11:12:57.897652+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":12,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.02955","citing_title":"Fast-dLLM++: Fr\\'{e}chet Profile Decoding for Faster Diffusion LLM Inference","ref_index":32,"is_internal_anchor":false},{"citing_arxiv_id":"2605.19262","citing_title":"Backdooring Masked Diffusion Language Models","ref_index":23,"is_internal_anchor":false},{"citing_arxiv_id":"2606.00295","citing_title":"Adaptive Order Policies for Masked Diffusion","ref_index":126,"is_internal_anchor":false},{"citing_arxiv_id":"2605.20813","citing_title":"PulseCol: Periodically Refreshed Column-Sparse Attention for Accelerating Diffusion Language Models","ref_index":2,"is_internal_anchor":false},{"citing_arxiv_id":"2605.19262","citing_title":"Backdooring Masked Diffusion Language Models","ref_index":23,"is_internal_anchor":false},{"citing_arxiv_id":"2605.16941","citing_title":"Roll Out and Roll Back: Diffusion LLMs are Their Own Efficiency Teachers","ref_index":28,"is_internal_anchor":false},{"citing_arxiv_id":"2604.08302","citing_title":"DMax: Aggressive Parallel Decoding for dLLMs","ref_index":8,"is_internal_anchor":false},{"citing_arxiv_id":"2510.04525","citing_title":"Demystifying MaskGIT Sampler and Beyond: Adaptive Order Selection in Masked Diffusion","ref_index":1,"is_internal_anchor":false},{"citing_arxiv_id":"2605.12522","citing_title":"Differences in Text Generated by Diffusion and Autoregressive Language Models","ref_index":3,"is_internal_anchor":false},{"citing_arxiv_id":"2605.10980","citing_title":"LEAP: Unlocking dLLM Parallelism via Lookahead Early-Convergence Token Detection","ref_index":3,"is_internal_anchor":false},{"citing_arxiv_id":"2604.08302","citing_title":"DMax: Aggressive Parallel Decoding for dLLMs","ref_index":8,"is_internal_anchor":false},{"citing_arxiv_id":"2604.17068","citing_title":"Stability-Weighted Decoding for Diffusion Language Models","ref_index":2,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/CZ6EL7JCVVXAZRVJ5KCH2KSXI7","json":"https://pith.science/pith/CZ6EL7JCVVXAZRVJ5KCH2KSXI7.json","graph_json":"https://pith.science/api/pith-number/CZ6EL7JCVVXAZRVJ5KCH2KSXI7/graph.json","events_json":"https://pith.science/api/pith-number/CZ6EL7JCVVXAZRVJ5KCH2KSXI7/events.json","paper":"https://pith.science/paper/CZ6EL7JC"},"agent_actions":{"view_html":"https://pith.science/pith/CZ6EL7JCVVXAZRVJ5KCH2KSXI7","download_json":"https://pith.science/pith/CZ6EL7JCVVXAZRVJ5KCH2KSXI7.json","view_paper":"https://pith.science/paper/CZ6EL7JC","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2505.24857&json=true","fetch_graph":"https://pith.science/api/pith-number/CZ6EL7JCVVXAZRVJ5KCH2KSXI7/graph.json","fetch_events":"https://pith.science/api/pith-number/CZ6EL7JCVVXAZRVJ5KCH2KSXI7/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/CZ6EL7JCVVXAZRVJ5KCH2KSXI7/action/timestamp_anchor","attest_storage":"https://pith.science/pith/CZ6EL7JCVVXAZRVJ5KCH2KSXI7/action/storage_attestation","attest_author":"https://pith.science/pith/CZ6EL7JCVVXAZRVJ5KCH2KSXI7/action/author_attestation","sign_citation":"https://pith.science/pith/CZ6EL7JCVVXAZRVJ5KCH2KSXI7/action/citation_signature","submit_replication":"https://pith.science/pith/CZ6EL7JCVVXAZRVJ5KCH2KSXI7/action/replication_record"}},"created_at":"2026-07-05T11:12:57.897652+00:00","updated_at":"2026-07-05T11:12:57.897652+00:00"}