{"paper":{"title":"Gradient-Discrepancy Acquisition for Pool-Based Active Learning","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"A gradient-discrepancy measure derived from a generalization bound serves as an effective acquisition criterion for pool-based active learning.","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Mohamadsadegh Khosravani, Sandra Zilles","submitted_at":"2026-05-04T13:56:09Z","abstract_excerpt":"The effectiveness of active learning hinges on the choice of the acquisition criterion by which a learning algorithm selects potentially informative data points whose label is subsequently queried. This paper proposes a novel gradient-based acquisition criterion, derived from a generalization bound introduced by Luo et al. (2022). This criterion can be applied in lieu of uncertainty measures in uncertainty sampling, or incorporated into diversity-based methods that consider the spread of sampled points in addition to the uncertainty of their labels. We provide a theoretical justification of th"},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"A novel gradient-based acquisition criterion, derived from a generalization bound introduced by Luo et al. (2022), can be applied in lieu of uncertainty measures in uncertainty sampling or incorporated into diversity-based methods.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"The generalization bound from Luo et al. (2022) can be directly leveraged to create an acquisition criterion that effectively identifies informative points beyond standard uncertainty or diversity measures.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"Introduces gradient-discrepancy acquisition criterion derived from Luo et al. (2022) generalization bound for active learning.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"A gradient-discrepancy measure derived from a generalization bound serves as an effective acquisition criterion for pool-based active learning.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"2d87dd6073d2c845df88133eb7e4ec6ca76f22f1b82bb47fd5f28b8ade1a5ff9"},"source":{"id":"2605.02609","kind":"arxiv","version":2},"verdict":{"id":"edb0d6bc-3d4e-4117-aeb1-bf310c0b81cb","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-19T17:15:07.399804Z","strongest_claim":"A novel gradient-based acquisition criterion, derived from a generalization bound introduced by Luo et al. (2022), can be applied in lieu of uncertainty measures in uncertainty sampling or incorporated into diversity-based methods.","one_line_summary":"Introduces gradient-discrepancy acquisition criterion derived from Luo et al. (2022) generalization bound for active learning.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"The generalization bound from Luo et al. (2022) can be directly leveraged to create an acquisition criterion that effectively identifies informative points beyond standard uncertainty or diversity measures.","pith_extraction_headline":"A gradient-discrepancy measure derived from a generalization bound serves as an effective acquisition criterion for pool-based active learning."},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2605.02609/integrity.json","findings":[],"available":true,"detectors_run":[{"name":"doi_compliance","ran_at":"2026-05-19T16:11:33.594657Z","status":"completed","version":"1.0.0","findings_count":0}],"snapshot_sha256":"71adfef172bc470f8fbac7b841813283c2e3af7868e89e29e4d5409541bceab0"},"references":{"count":15,"sample":[{"doi":"","year":2025,"title":"Accessed 2025-12-15","work_id":"b2810175-f342-4116-b0aa-575e5a7edb17","ref_index":1,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2025,"title":"Accessed 2025-12-15","work_id":"8883bb86-7834-4500-bcba-6f773f420fc5","ref_index":2,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2025,"title":"Accessed 2025-12-15","work_id":"844b4e8e-c89a-4d22-8e65-f549b845c330","ref_index":3,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2025,"title":"Accessed 2025-12-15","work_id":"4cecec85-b816-4a8f-8517-3692c0dcd326","ref_index":4,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"10.1007/s101070100263","year":null,"title":"Benchmarking Optimization Software with Performance Profiles","work_id":"c78cf4ec-cc7f-4ba9-8d51-1a183cdfb3a0","ref_index":5,"cited_arxiv_id":"","is_internal_anchor":false}],"resolved_work":15,"snapshot_sha256":"763070a7c44e1d9ea368ef28702518f736e114a45eac96c5bd9e0f8e6319d5e8","internal_anchors":2},"formal_canon":{"evidence_count":2,"snapshot_sha256":"dafc2b362a5eb35f7cbbda9555f61b642d2a9f38c5b2218a5d79d03bb2152160"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"}