{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2023:DFVYRA3FMQDNNDIPZKY3XMSWEF","short_pith_number":"pith:DFVYRA3F","canonical_record":{"source":{"id":"2310.10944","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2023-10-17T02:42:34Z","cross_cats_sorted":[],"title_canon_sha256":"73fc7dd99a27665eeff37290f8a68b3de406ef7765a64199e94cc2995f064709","abstract_canon_sha256":"18191522f0f96fe397d05e9dec7f3c0838525a3a57473b2d2e2033dbe7a3f63f"},"schema_version":"1.0"},"canonical_sha256":"196b8883656406d68d0fcab1bbb256214543e1f07fc2d76ff4e86352cfc63b85","source":{"kind":"arxiv","id":"2310.10944","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2310.10944","created_at":"2026-07-05T07:01:29Z"},{"alias_kind":"arxiv_version","alias_value":"2310.10944v1","created_at":"2026-07-05T07:01:29Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2310.10944","created_at":"2026-07-05T07:01:29Z"},{"alias_kind":"pith_short_12","alias_value":"DFVYRA3FMQDN","created_at":"2026-07-05T07:01:29Z"},{"alias_kind":"pith_short_16","alias_value":"DFVYRA3FMQDNNDIP","created_at":"2026-07-05T07:01:29Z"},{"alias_kind":"pith_short_8","alias_value":"DFVYRA3F","created_at":"2026-07-05T07:01:29Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2023:DFVYRA3FMQDNNDIPZKY3XMSWEF","target":"record","payload":{"canonical_record":{"source":{"id":"2310.10944","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2023-10-17T02:42:34Z","cross_cats_sorted":[],"title_canon_sha256":"73fc7dd99a27665eeff37290f8a68b3de406ef7765a64199e94cc2995f064709","abstract_canon_sha256":"18191522f0f96fe397d05e9dec7f3c0838525a3a57473b2d2e2033dbe7a3f63f"},"schema_version":"1.0"},"canonical_sha256":"196b8883656406d68d0fcab1bbb256214543e1f07fc2d76ff4e86352cfc63b85","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T07:01:29.864091Z","signature_b64":"ejg94IhSA2WUHzI/kqjcQVs1cGVNU9JMTRqFmAlVGX/lJyqRxCvJlsAz92IL0IQmUm7x/FlN8OLLylI0/ME8Dw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"196b8883656406d68d0fcab1bbb256214543e1f07fc2d76ff4e86352cfc63b85","last_reissued_at":"2026-07-05T07:01:29.863543Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T07:01:29.863543Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2310.10944","source_version":1,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T07:01:29Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"JarS3fB4Djt6O626AcsGwOT81TKi+w2K8Zt29t4t8G62SSGEA890DmSVCEcmReXLCUalzo7z9lGhEmQ6LEelDA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-09T12:08:14.907758Z"},"content_sha256":"ee85d9eed29a4726727cca77450b3f8f515aa6c928de7d8115a09a5edb5f37e3","schema_version":"1.0","event_id":"sha256:ee85d9eed29a4726727cca77450b3f8f515aa6c928de7d8115a09a5edb5f37e3"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2023:DFVYRA3FMQDNNDIPZKY3XMSWEF","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"TEQ: Trainable Equivalent Transformation for Quantization of LLMs","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Haihao Shen, Kaokao Lv, Wenhua Cheng, Yiyang Cai","submitted_at":"2023-10-17T02:42:34Z","abstract_excerpt":"As large language models (LLMs) become more prevalent, there is a growing need for new and improved quantization methods that can meet the computationalast layer demands of these modern architectures while maintaining the accuracy. In this paper, we present TEQ, a trainable equivalent transformation that preserves the FP32 precision of the model output while taking advantage of low-precision quantization, especially 3 and 4 bits weight-only quantization. The training process is lightweight, requiring only 1K steps and fewer than 0.1 percent of the original model's trainable parameters. Further"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2310.10944","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/2310.10944/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":"2026-07-05T07:01:29Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"fYfqJA0DSOD5TiUUJwIfoQVuKj2nET1LJBJM9bgf7cy5+AGNCQGcOgWycGm507yCL2INs9b1F2KusJJ+LUJJBg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-09T12:08:14.908250Z"},"content_sha256":"ce03a0a36080426fdb3daf2795d80c9e0a7177f2044e07e63534e60dbb73d438","schema_version":"1.0","event_id":"sha256:ce03a0a36080426fdb3daf2795d80c9e0a7177f2044e07e63534e60dbb73d438"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/DFVYRA3FMQDNNDIPZKY3XMSWEF/bundle.json","state_url":"https://pith.science/pith/DFVYRA3FMQDNNDIPZKY3XMSWEF/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/DFVYRA3FMQDNNDIPZKY3XMSWEF/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":"2026-08-09T12:08:14Z","links":{"resolver":"https://pith.science/pith/DFVYRA3FMQDNNDIPZKY3XMSWEF","bundle":"https://pith.science/pith/DFVYRA3FMQDNNDIPZKY3XMSWEF/bundle.json","state":"https://pith.science/pith/DFVYRA3FMQDNNDIPZKY3XMSWEF/state.json","well_known_bundle":"https://pith.science/.well-known/pith/DFVYRA3FMQDNNDIPZKY3XMSWEF/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2023:DFVYRA3FMQDNNDIPZKY3XMSWEF","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":"18191522f0f96fe397d05e9dec7f3c0838525a3a57473b2d2e2033dbe7a3f63f","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2023-10-17T02:42:34Z","title_canon_sha256":"73fc7dd99a27665eeff37290f8a68b3de406ef7765a64199e94cc2995f064709"},"schema_version":"1.0","source":{"id":"2310.10944","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2310.10944","created_at":"2026-07-05T07:01:29Z"},{"alias_kind":"arxiv_version","alias_value":"2310.10944v1","created_at":"2026-07-05T07:01:29Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2310.10944","created_at":"2026-07-05T07:01:29Z"},{"alias_kind":"pith_short_12","alias_value":"DFVYRA3FMQDN","created_at":"2026-07-05T07:01:29Z"},{"alias_kind":"pith_short_16","alias_value":"DFVYRA3FMQDNNDIP","created_at":"2026-07-05T07:01:29Z"},{"alias_kind":"pith_short_8","alias_value":"DFVYRA3F","created_at":"2026-07-05T07:01:29Z"}],"graph_snapshots":[{"event_id":"sha256:ce03a0a36080426fdb3daf2795d80c9e0a7177f2044e07e63534e60dbb73d438","target":"graph","created_at":"2026-07-05T07:01:29Z","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/2310.10944/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"As large language models (LLMs) become more prevalent, there is a growing need for new and improved quantization methods that can meet the computationalast layer demands of these modern architectures while maintaining the accuracy. In this paper, we present TEQ, a trainable equivalent transformation that preserves the FP32 precision of the model output while taking advantage of low-precision quantization, especially 3 and 4 bits weight-only quantization. The training process is lightweight, requiring only 1K steps and fewer than 0.1 percent of the original model's trainable parameters. Further","authors_text":"Haihao Shen, Kaokao Lv, Wenhua Cheng, Yiyang Cai","cross_cats":[],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2023-10-17T02:42:34Z","title":"TEQ: Trainable Equivalent Transformation for Quantization of LLMs"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2310.10944","kind":"arxiv","version":1},"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:ee85d9eed29a4726727cca77450b3f8f515aa6c928de7d8115a09a5edb5f37e3","target":"record","created_at":"2026-07-05T07:01:29Z","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":"18191522f0f96fe397d05e9dec7f3c0838525a3a57473b2d2e2033dbe7a3f63f","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2023-10-17T02:42:34Z","title_canon_sha256":"73fc7dd99a27665eeff37290f8a68b3de406ef7765a64199e94cc2995f064709"},"schema_version":"1.0","source":{"id":"2310.10944","kind":"arxiv","version":1}},"canonical_sha256":"196b8883656406d68d0fcab1bbb256214543e1f07fc2d76ff4e86352cfc63b85","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"196b8883656406d68d0fcab1bbb256214543e1f07fc2d76ff4e86352cfc63b85","first_computed_at":"2026-07-05T07:01:29.863543Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T07:01:29.863543Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"ejg94IhSA2WUHzI/kqjcQVs1cGVNU9JMTRqFmAlVGX/lJyqRxCvJlsAz92IL0IQmUm7x/FlN8OLLylI0/ME8Dw==","signature_status":"signed_v1","signed_at":"2026-07-05T07:01:29.864091Z","signed_message":"canonical_sha256_bytes"},"source_id":"2310.10944","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:ee85d9eed29a4726727cca77450b3f8f515aa6c928de7d8115a09a5edb5f37e3","sha256:ce03a0a36080426fdb3daf2795d80c9e0a7177f2044e07e63534e60dbb73d438"],"state_sha256":"ae4f526ccc2c41da75970a1ed3d1a611e0b273c77ae44e531d0dc797a90bf491"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"9TXi8FRPmiEM/OEpEhFVY3W/R6FPp//1gG0hRBB4j64yXsmAEg5Trr0pc7ZjQT+RNwIIzuqni8Ss/bJRgmOsCw==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-09T12:08:14.912109Z","bundle_sha256":"0ddbd66a1b8b69fa42394f45baacfc86eda9ab9e91e82fef6a986bb345675ed3"}}