{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2019:U3XSQL4A7EIOMLL2BCD3L2QX23","short_pith_number":"pith:U3XSQL4A","canonical_record":{"source":{"id":"1909.13271","kind":"arxiv","version":3},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2019-09-29T12:41:46Z","cross_cats_sorted":["cs.AR","stat.ML"],"title_canon_sha256":"004d4aa2b14c0b22f2b547dd70e46daa7e54c824c8ffe5d65ae13b2a9e20ba09","abstract_canon_sha256":"bb38775d96dbe1752239b730ff0790e48d4c9cf44f39b6365bf13738aa8cb6dd"},"schema_version":"1.0"},"canonical_sha256":"a6ef282f80f910e62d7a0887b5ea17d6e7f8d9acaa9f9aa99fa0c396bf9f9c6b","source":{"kind":"arxiv","id":"1909.13271","version":3},"source_aliases":[{"alias_kind":"arxiv","alias_value":"1909.13271","created_at":"2026-07-05T00:39:38Z"},{"alias_kind":"arxiv_version","alias_value":"1909.13271v3","created_at":"2026-07-05T00:39:38Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1909.13271","created_at":"2026-07-05T00:39:38Z"},{"alias_kind":"pith_short_12","alias_value":"U3XSQL4A7EIO","created_at":"2026-07-05T00:39:38Z"},{"alias_kind":"pith_short_16","alias_value":"U3XSQL4A7EIOMLL2","created_at":"2026-07-05T00:39:38Z"},{"alias_kind":"pith_short_8","alias_value":"U3XSQL4A","created_at":"2026-07-05T00:39:38Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2019:U3XSQL4A7EIOMLL2BCD3L2QX23","target":"record","payload":{"canonical_record":{"source":{"id":"1909.13271","kind":"arxiv","version":3},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2019-09-29T12:41:46Z","cross_cats_sorted":["cs.AR","stat.ML"],"title_canon_sha256":"004d4aa2b14c0b22f2b547dd70e46daa7e54c824c8ffe5d65ae13b2a9e20ba09","abstract_canon_sha256":"bb38775d96dbe1752239b730ff0790e48d4c9cf44f39b6365bf13738aa8cb6dd"},"schema_version":"1.0"},"canonical_sha256":"a6ef282f80f910e62d7a0887b5ea17d6e7f8d9acaa9f9aa99fa0c396bf9f9c6b","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T00:39:38.742566Z","signature_b64":"mM7BJ7/LrmTq58pV6Fy2b0JMNquotFOqyqYXK00NCFoKZXPmwWUINr7cQNROfUwIrA/H5O0x8uaLejmKG+VPCw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"a6ef282f80f910e62d7a0887b5ea17d6e7f8d9acaa9f9aa99fa0c396bf9f9c6b","last_reissued_at":"2026-07-05T00:39:38.742072Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T00:39:38.742072Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"1909.13271","source_version":3,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T00:39:38Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"q4mOunNjPY5XQArAzUsrxDSM9msgp6Y6Qkqo+w9gW9FssBQ8elE5fbW0aWtlcqtuCHTnggE9OgG86W//xc3rCQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-22T01:35:14.624085Z"},"content_sha256":"06b77cb00ded37593ed9cc54589ccf1eb3520a484fb368368d584ca18002b893","schema_version":"1.0","event_id":"sha256:06b77cb00ded37593ed9cc54589ccf1eb3520a484fb368368d584ca18002b893"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2019:U3XSQL4A7EIOMLL2BCD3L2QX23","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"AdaptivFloat: A Floating-point based Data Type for Resilient Deep Learning Inference","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AR","stat.ML"],"primary_cat":"cs.LG","authors_text":"Alexander Rush, David Brooks, En-Yu Yang, Gu-Yeon Wei, Thierry Tambe, Vijay Janapa Reddi, Yuntian Deng, Zishen Wan","submitted_at":"2019-09-29T12:41:46Z","abstract_excerpt":"Conventional hardware-friendly quantization methods, such as fixed-point or integer, tend to perform poorly at very low word sizes as their shrinking dynamic ranges cannot adequately capture the wide data distributions commonly seen in sequence transduction models. We present AdaptivFloat, a floating-point inspired number representation format for deep learning that dynamically maximizes and optimally clips its available dynamic range, at a layer granularity, in order to create faithful encoding of neural network parameters. AdaptivFloat consistently produces higher inference accuracies compar"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"1909.13271","kind":"arxiv","version":3},"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/1909.13271/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"},"verdict_id":null},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T00:39:38Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"MKcpkLchn+Gh3806xZOGO4aNVi0jPRVJUv8w9+8ozcuyxTjRRHOyTPJYcMaM6pBWkTzQO+8XD578MJQhmQfBCw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-22T01:35:14.624578Z"},"content_sha256":"3ebd079ebe59d08d6e48f73ce57913c41c4497c6e779cfeab04e4d1094c36a1d","schema_version":"1.0","event_id":"sha256:3ebd079ebe59d08d6e48f73ce57913c41c4497c6e779cfeab04e4d1094c36a1d"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/U3XSQL4A7EIOMLL2BCD3L2QX23/bundle.json","state_url":"https://pith.science/pith/U3XSQL4A7EIOMLL2BCD3L2QX23/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/U3XSQL4A7EIOMLL2BCD3L2QX23/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":"2026-08-22T01:35:14Z","links":{"resolver":"https://pith.science/pith/U3XSQL4A7EIOMLL2BCD3L2QX23","bundle":"https://pith.science/pith/U3XSQL4A7EIOMLL2BCD3L2QX23/bundle.json","state":"https://pith.science/pith/U3XSQL4A7EIOMLL2BCD3L2QX23/state.json","well_known_bundle":"https://pith.science/.well-known/pith/U3XSQL4A7EIOMLL2BCD3L2QX23/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2019:U3XSQL4A7EIOMLL2BCD3L2QX23","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":"bb38775d96dbe1752239b730ff0790e48d4c9cf44f39b6365bf13738aa8cb6dd","cross_cats_sorted":["cs.AR","stat.ML"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2019-09-29T12:41:46Z","title_canon_sha256":"004d4aa2b14c0b22f2b547dd70e46daa7e54c824c8ffe5d65ae13b2a9e20ba09"},"schema_version":"1.0","source":{"id":"1909.13271","kind":"arxiv","version":3}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"1909.13271","created_at":"2026-07-05T00:39:38Z"},{"alias_kind":"arxiv_version","alias_value":"1909.13271v3","created_at":"2026-07-05T00:39:38Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1909.13271","created_at":"2026-07-05T00:39:38Z"},{"alias_kind":"pith_short_12","alias_value":"U3XSQL4A7EIO","created_at":"2026-07-05T00:39:38Z"},{"alias_kind":"pith_short_16","alias_value":"U3XSQL4A7EIOMLL2","created_at":"2026-07-05T00:39:38Z"},{"alias_kind":"pith_short_8","alias_value":"U3XSQL4A","created_at":"2026-07-05T00:39:38Z"}],"graph_snapshots":[{"event_id":"sha256:3ebd079ebe59d08d6e48f73ce57913c41c4497c6e779cfeab04e4d1094c36a1d","target":"graph","created_at":"2026-07-05T00:39:38Z","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":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"integrity":{"available":true,"clean":true,"detectors_run":[],"endpoint":"/pith/1909.13271/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Conventional hardware-friendly quantization methods, such as fixed-point or integer, tend to perform poorly at very low word sizes as their shrinking dynamic ranges cannot adequately capture the wide data distributions commonly seen in sequence transduction models. We present AdaptivFloat, a floating-point inspired number representation format for deep learning that dynamically maximizes and optimally clips its available dynamic range, at a layer granularity, in order to create faithful encoding of neural network parameters. AdaptivFloat consistently produces higher inference accuracies compar","authors_text":"Alexander Rush, David Brooks, En-Yu Yang, Gu-Yeon Wei, Thierry Tambe, Vijay Janapa Reddi, Yuntian Deng, Zishen Wan","cross_cats":["cs.AR","stat.ML"],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2019-09-29T12:41:46Z","title":"AdaptivFloat: A Floating-point based Data Type for Resilient Deep Learning Inference"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"1909.13271","kind":"arxiv","version":3},"verdict":{"created_at":null,"id":null,"model_set":{},"one_line_summary":"","pipeline_version":null,"pith_extraction_headline":"","strongest_claim":"","weakest_assumption":""}},"verdict_id":null}}],"author_attestations":[],"timestamp_anchors":[],"storage_attestations":[],"citation_signatures":[],"replication_records":[],"corrections":[],"mirror_hints":[],"record_created":{"event_id":"sha256:06b77cb00ded37593ed9cc54589ccf1eb3520a484fb368368d584ca18002b893","target":"record","created_at":"2026-07-05T00:39:38Z","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":"bb38775d96dbe1752239b730ff0790e48d4c9cf44f39b6365bf13738aa8cb6dd","cross_cats_sorted":["cs.AR","stat.ML"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2019-09-29T12:41:46Z","title_canon_sha256":"004d4aa2b14c0b22f2b547dd70e46daa7e54c824c8ffe5d65ae13b2a9e20ba09"},"schema_version":"1.0","source":{"id":"1909.13271","kind":"arxiv","version":3}},"canonical_sha256":"a6ef282f80f910e62d7a0887b5ea17d6e7f8d9acaa9f9aa99fa0c396bf9f9c6b","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"a6ef282f80f910e62d7a0887b5ea17d6e7f8d9acaa9f9aa99fa0c396bf9f9c6b","first_computed_at":"2026-07-05T00:39:38.742072Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T00:39:38.742072Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"mM7BJ7/LrmTq58pV6Fy2b0JMNquotFOqyqYXK00NCFoKZXPmwWUINr7cQNROfUwIrA/H5O0x8uaLejmKG+VPCw==","signature_status":"signed_v1","signed_at":"2026-07-05T00:39:38.742566Z","signed_message":"canonical_sha256_bytes"},"source_id":"1909.13271","source_kind":"arxiv","source_version":3}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:06b77cb00ded37593ed9cc54589ccf1eb3520a484fb368368d584ca18002b893","sha256:3ebd079ebe59d08d6e48f73ce57913c41c4497c6e779cfeab04e4d1094c36a1d"],"state_sha256":"54ba9757791774a421de220de2040aa7d2c2590e8c3e960137a03d50d08961ca"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"OWV1npGC+P+RdrdHVISm18PV6cwpRpqTZz5Hgx0eZlF5hqGsopSNxfZkXcyS6d+MSSnNbgMXftag7gawg5p2CA==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-22T01:35:14.628481Z","bundle_sha256":"cdb976e30d5fec9b0acf4e81b2f47027292e2fc59f07f5f9e946d615dd8e7ad8"}}