{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:LA7SLZTYVFWDLRXWPEOGRTSUMK","short_pith_number":"pith:LA7SLZTY","schema_version":"1.0","canonical_sha256":"583f25e678a96c35c6f6791c68ce5462b0409a1c1da364a88506c9c0ce219983","source":{"kind":"arxiv","id":"2508.16700","version":2},"attestation_state":"computed","paper":{"title":"GPT-OSS-20B: A Comprehensive Deployment-Centric Analysis of OpenAI's Open-Weight Mixture of Experts Model","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.DC","cs.PF"],"primary_cat":"cs.AR","authors_text":"Deepak Kumar, Divakar Yadav, Yash Patel","submitted_at":"2025-08-22T03:37:27Z","abstract_excerpt":"We present a single-GPU (H100, bf16) evaluation of GPT-OSS-20B (Mixture-of-Experts; 20.9B total, approx. 3.61B active) against dense baselines Qwen3-32B and Yi-34B across multiple dimensions. We measure true time-to-first-token (TTFT), full-decode throughput (TPOT), end-to-end latency percentiles, peak VRAM with past key values (PKV) held, and energy via a consistent nvidia-smi-based sampler. At a 2048-token context with 64-token decode, GPT-OSS-20B delivers higher decode throughput and tokens per Joule than dense baselines Qwen3-32B and Yi-34B, while substantially reducing peak VRAM and energ"},"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":"2508.16700","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AR","submitted_at":"2025-08-22T03:37:27Z","cross_cats_sorted":["cs.AI","cs.DC","cs.PF"],"title_canon_sha256":"2e94c346a462b09182f6b0d2a24303c4be04f7c95029f180fa449f8bd30c81c0","abstract_canon_sha256":"21c5afab51333b615e603ff26df759455ea1d92f995577aa271c303685f6d53f"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T12:02:06.641772Z","signature_b64":"TKyNybaX6kh9O5RYn6uu1KT6FAeU2onEki3Pfu9oCKvCEd34hq0OSCsfse+ZlNZXU1/7i6q5qc223XqqZnydCg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"583f25e678a96c35c6f6791c68ce5462b0409a1c1da364a88506c9c0ce219983","last_reissued_at":"2026-07-05T12:02:06.641102Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T12:02:06.641102Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"GPT-OSS-20B: A Comprehensive Deployment-Centric Analysis of OpenAI's Open-Weight Mixture of Experts Model","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.DC","cs.PF"],"primary_cat":"cs.AR","authors_text":"Deepak Kumar, Divakar Yadav, Yash Patel","submitted_at":"2025-08-22T03:37:27Z","abstract_excerpt":"We present a single-GPU (H100, bf16) evaluation of GPT-OSS-20B (Mixture-of-Experts; 20.9B total, approx. 3.61B active) against dense baselines Qwen3-32B and Yi-34B across multiple dimensions. We measure true time-to-first-token (TTFT), full-decode throughput (TPOT), end-to-end latency percentiles, peak VRAM with past key values (PKV) held, and energy via a consistent nvidia-smi-based sampler. At a 2048-token context with 64-token decode, GPT-OSS-20B delivers higher decode throughput and tokens per Joule than dense baselines Qwen3-32B and Yi-34B, while substantially reducing peak VRAM and energ"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2508.16700","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/2508.16700/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":"2508.16700","created_at":"2026-07-05T12:02:06.641183+00:00"},{"alias_kind":"arxiv_version","alias_value":"2508.16700v2","created_at":"2026-07-05T12:02:06.641183+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2508.16700","created_at":"2026-07-05T12:02:06.641183+00:00"},{"alias_kind":"pith_short_12","alias_value":"LA7SLZTYVFWD","created_at":"2026-07-05T12:02:06.641183+00:00"},{"alias_kind":"pith_short_16","alias_value":"LA7SLZTYVFWDLRXW","created_at":"2026-07-05T12:02:06.641183+00:00"},{"alias_kind":"pith_short_8","alias_value":"LA7SLZTY","created_at":"2026-07-05T12:02:06.641183+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2512.12794","citing_title":"A Rule-Aware Prompt Framework for Structured Numeric Reasoning in Cyber-Physical Systems","ref_index":17,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/LA7SLZTYVFWDLRXWPEOGRTSUMK","json":"https://pith.science/pith/LA7SLZTYVFWDLRXWPEOGRTSUMK.json","graph_json":"https://pith.science/api/pith-number/LA7SLZTYVFWDLRXWPEOGRTSUMK/graph.json","events_json":"https://pith.science/api/pith-number/LA7SLZTYVFWDLRXWPEOGRTSUMK/events.json","paper":"https://pith.science/paper/LA7SLZTY"},"agent_actions":{"view_html":"https://pith.science/pith/LA7SLZTYVFWDLRXWPEOGRTSUMK","download_json":"https://pith.science/pith/LA7SLZTYVFWDLRXWPEOGRTSUMK.json","view_paper":"https://pith.science/paper/LA7SLZTY","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2508.16700&json=true","fetch_graph":"https://pith.science/api/pith-number/LA7SLZTYVFWDLRXWPEOGRTSUMK/graph.json","fetch_events":"https://pith.science/api/pith-number/LA7SLZTYVFWDLRXWPEOGRTSUMK/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/LA7SLZTYVFWDLRXWPEOGRTSUMK/action/timestamp_anchor","attest_storage":"https://pith.science/pith/LA7SLZTYVFWDLRXWPEOGRTSUMK/action/storage_attestation","attest_author":"https://pith.science/pith/LA7SLZTYVFWDLRXWPEOGRTSUMK/action/author_attestation","sign_citation":"https://pith.science/pith/LA7SLZTYVFWDLRXWPEOGRTSUMK/action/citation_signature","submit_replication":"https://pith.science/pith/LA7SLZTYVFWDLRXWPEOGRTSUMK/action/replication_record"}},"created_at":"2026-07-05T12:02:06.641183+00:00","updated_at":"2026-07-05T12:02:06.641183+00:00"}