{"paper":{"title":"A New Class of Geometric Analog Error Correction Codes for Crossbar Based In-Memory Computing","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"A geometric analysis characterizes m-height profiles of a family of analog codes to handle multiple outliers in resistive crossbar computing.","cross_cats":["math.IT"],"primary_cat":"cs.IT","authors_text":"Anxiao Jiang, Changcheng Yuan, Paul H. Siegel, Ron M. Roth, Ziyuan Zhu","submitted_at":"2026-03-04T04:51:30Z","abstract_excerpt":"Analog error correction codes have been proposed for analog in-memory computing on resistive crossbars, which can accelerate vector-matrix multiplication for machine learning. Unlike traditional communication or storage channels, this setting involves a mixed noise model with small perturbations and outlier errors. A number of analog codes have been proposed for handling a single outlier, and several constructions have also been developed to address multiple outliers. However, the set of available code families remains limited, covering only a narrow range of code lengths and dimensions. In th"},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"we study a recently proposed family of geometric codes capable of handling multiple outliers, and develop a geometric analysis that characterizes their m-height profiles.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"That the recently proposed geometric codes are indeed capable of handling multiple outliers under the mixed noise model of resistive crossbars, and that a geometric analysis of m-height profiles will meaningfully expand the usable code families.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"The authors develop a geometric analysis that characterizes the m-height profiles of geometric codes capable of correcting multiple outliers in analog in-memory computing.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"A geometric analysis characterizes m-height profiles of a family of analog codes to handle multiple outliers in resistive crossbar computing.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"56bb9d167158a40544e4bc28b26a14c230c0a1cb9066adb890030dac92f369d0"},"source":{"id":"2603.03723","kind":"arxiv","version":4},"verdict":{"id":"efd5002c-8336-4789-bba6-0ac3276faf4d","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-15T17:10:38.530968Z","strongest_claim":"we study a recently proposed family of geometric codes capable of handling multiple outliers, and develop a geometric analysis that characterizes their m-height profiles.","one_line_summary":"The authors develop a geometric analysis that characterizes the m-height profiles of geometric codes capable of correcting multiple outliers in analog in-memory computing.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"That the recently proposed geometric codes are indeed capable of handling multiple outliers under the mixed noise model of resistive crossbars, and that a geometric analysis of m-height profiles will meaningfully expand the usable code families.","pith_extraction_headline":"A geometric analysis characterizes m-height profiles of a family of analog codes to handle multiple outliers in resistive crossbar computing."},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2603.03723/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":2,"snapshot_sha256":"0059c9dc31691210c97a192821386564cedba8264f85801a5203ce185e48e1eb"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"}