{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2026:C67XSDJHSXOMZV4ZPPXFT74GIS","short_pith_number":"pith:C67XSDJH","canonical_record":{"source":{"id":"2604.24810","kind":"arxiv","version":3},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-04-27T08:51:44Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"8298ff77409caa6bd2038e49aadafb2e4053ae1733440fa7aee06fd2fd961f44","abstract_canon_sha256":"d3ef120d0fa5b8e02c510662b287655029ea01cca648c04094e654637d5a3f73"},"schema_version":"1.0"},"canonical_sha256":"17bf790d2795dcccd7997bee59ff86449096c9bf46c34885e3d899c74d122dcd","source":{"kind":"arxiv","id":"2604.24810","version":3},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2604.24810","created_at":"2026-05-25T02:01:21Z"},{"alias_kind":"arxiv_version","alias_value":"2604.24810v3","created_at":"2026-05-25T02:01:21Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2604.24810","created_at":"2026-05-25T02:01:21Z"},{"alias_kind":"pith_short_12","alias_value":"C67XSDJHSXOM","created_at":"2026-05-25T02:01:21Z"},{"alias_kind":"pith_short_16","alias_value":"C67XSDJHSXOMZV4Z","created_at":"2026-05-25T02:01:21Z"},{"alias_kind":"pith_short_8","alias_value":"C67XSDJH","created_at":"2026-05-25T02:01:21Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2026:C67XSDJHSXOMZV4ZPPXFT74GIS","target":"record","payload":{"canonical_record":{"source":{"id":"2604.24810","kind":"arxiv","version":3},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-04-27T08:51:44Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"8298ff77409caa6bd2038e49aadafb2e4053ae1733440fa7aee06fd2fd961f44","abstract_canon_sha256":"d3ef120d0fa5b8e02c510662b287655029ea01cca648c04094e654637d5a3f73"},"schema_version":"1.0"},"canonical_sha256":"17bf790d2795dcccd7997bee59ff86449096c9bf46c34885e3d899c74d122dcd","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-25T02:01:21.158845Z","signature_b64":"aRt4MkFmn7xz8YraLho740nyU15LasqTv2PUdGRXa7BGWPYkNMbAIJ9Om9WSQC0Uh8BzOBBVOpek40nnvAh8Dg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"17bf790d2795dcccd7997bee59ff86449096c9bf46c34885e3d899c74d122dcd","last_reissued_at":"2026-05-25T02:01:21.158039Z","signature_status":"signed_v1","first_computed_at":"2026-05-25T02:01:21.158039Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2604.24810","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":"2026-05-25T02:01:21Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"IpKS/VMNxh+WvPHcI49s/gLu3aP4jrZpTZkXj3XR3+qk1xN9GXjHdrnL2ehk1oW/S+mwm5B7WSyx4UpnXOENCA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-05-27T20:15:50.807998Z"},"content_sha256":"8b2cb1cb2fe3d470e55f74f8d709035c4dc4ea3beee937067cd799382200cda5","schema_version":"1.0","event_id":"sha256:8b2cb1cb2fe3d470e55f74f8d709035c4dc4ea3beee937067cd799382200cda5"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2026:C67XSDJHSXOMZV4ZPPXFT74GIS","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"A Comparative Analysis on the Performance of Upper Confidence Bound Algorithms in Adaptive Deep Neural Networks","license":"http://creativecommons.org/licenses/by/4.0/","headline":"Multiple Upper Confidence Bound strategies achieve sub-linear regret in Adaptive Deep Neural Networks and improve accuracy-energy and accuracy-latency trade-offs.","cross_cats":["cs.AI"],"primary_cat":"cs.LG","authors_text":"Grigorios Papanikolaou, Ioannis Kontopoulos, Konstantinos Tserpes","submitted_at":"2026-04-27T08:51:44Z","abstract_excerpt":"Edge computing environments impose strict constraints on energy consumption and latency, making the deployment of deep neural networks a significant challenge. Therefore, smart and adaptive inference strategies that dynamically balance computational cost or latency with predictive accuracy are critical in edge computing scenarios. In this work, we build on Adaptive Deep Neural Networks (ADNNs) that employ the Multi-Armed Bandit (MAB) framework. Current literature leverages the first version of the Upper Confidence Bound (UCB1) strategy to dynamically select the optimal confidence threshold, en"},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"Experimental results demonstrate that all strategies achieve sub-linear cumulative regret, with UCB-Bayes converging the fastest, followed by UCB-Tuned and UCB-V. Finally, UCB-V and UCB-Tuned dominate the Pareto Frontiers of accuracy-latency and accuracy-energy trade-offs.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"The assumption that the multi-armed bandit reward distributions remain stationary across inference steps and that the chosen datasets and models adequately represent real edge-device conditions with varying input distributions and hardware constraints.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"UCB-V and UCB-Tuned dominate accuracy-energy and accuracy-latency trade-offs while all tested UCB strategies achieve sub-linear regret in adaptive DNN early-exit experiments on CIFAR datasets.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"Multiple Upper Confidence Bound strategies achieve sub-linear regret in Adaptive Deep Neural Networks and improve accuracy-energy and accuracy-latency trade-offs.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"3a540bb6895f01179888b83ca0fcd8977fc5e2e0f0cb31d6ed1bb22256a5ee1e"},"source":{"id":"2604.24810","kind":"arxiv","version":3},"verdict":{"id":"96c191fb-8275-4f23-b425-b889be43d580","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-08T04:20:33.911517Z","strongest_claim":"Experimental results demonstrate that all strategies achieve sub-linear cumulative regret, with UCB-Bayes converging the fastest, followed by UCB-Tuned and UCB-V. Finally, UCB-V and UCB-Tuned dominate the Pareto Frontiers of accuracy-latency and accuracy-energy trade-offs.","one_line_summary":"UCB-V and UCB-Tuned dominate accuracy-energy and accuracy-latency trade-offs while all tested UCB strategies achieve sub-linear regret in adaptive DNN early-exit experiments on CIFAR datasets.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"The assumption that the multi-armed bandit reward distributions remain stationary across inference steps and that the chosen datasets and models adequately represent real edge-device conditions with varying input distributions and hardware constraints.","pith_extraction_headline":"Multiple Upper Confidence Bound strategies achieve sub-linear regret in Adaptive Deep Neural Networks and improve accuracy-energy and accuracy-latency trade-offs."},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2604.24810/integrity.json","findings":[],"available":true,"detectors_run":[{"name":"ai_meta_artifact","ran_at":"2026-05-21T07:36:22.860224Z","status":"completed","version":"1.0.0","findings_count":0},{"name":"doi_compliance","ran_at":"2026-05-19T22:23:14.663615Z","status":"completed","version":"1.0.0","findings_count":0}],"snapshot_sha256":"a119f36626d20b77ea381903fa106606cdb816b71db33090b17509c776b9f701"},"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":"96c191fb-8275-4f23-b425-b889be43d580"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-05-25T02:01:21Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"ljd0+5aAWf+pzmaDf76YvvTdoAOGxXgXseD5mWrNlUP82MNBAhcN/8561caWA9DMgC0bisfQCnXr9aaRxci0Aw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-05-27T20:15:50.808931Z"},"content_sha256":"dab93c22455c026eda5dd4fb10f3f63301b4bce6d03507c95d7106d7913d843f","schema_version":"1.0","event_id":"sha256:dab93c22455c026eda5dd4fb10f3f63301b4bce6d03507c95d7106d7913d843f"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/C67XSDJHSXOMZV4ZPPXFT74GIS/bundle.json","state_url":"https://pith.science/pith/C67XSDJHSXOMZV4ZPPXFT74GIS/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/C67XSDJHSXOMZV4ZPPXFT74GIS/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":"2026-05-27T20:15:50Z","links":{"resolver":"https://pith.science/pith/C67XSDJHSXOMZV4ZPPXFT74GIS","bundle":"https://pith.science/pith/C67XSDJHSXOMZV4ZPPXFT74GIS/bundle.json","state":"https://pith.science/pith/C67XSDJHSXOMZV4ZPPXFT74GIS/state.json","well_known_bundle":"https://pith.science/.well-known/pith/C67XSDJHSXOMZV4ZPPXFT74GIS/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:C67XSDJHSXOMZV4ZPPXFT74GIS","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":"d3ef120d0fa5b8e02c510662b287655029ea01cca648c04094e654637d5a3f73","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-04-27T08:51:44Z","title_canon_sha256":"8298ff77409caa6bd2038e49aadafb2e4053ae1733440fa7aee06fd2fd961f44"},"schema_version":"1.0","source":{"id":"2604.24810","kind":"arxiv","version":3}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2604.24810","created_at":"2026-05-25T02:01:21Z"},{"alias_kind":"arxiv_version","alias_value":"2604.24810v3","created_at":"2026-05-25T02:01:21Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2604.24810","created_at":"2026-05-25T02:01:21Z"},{"alias_kind":"pith_short_12","alias_value":"C67XSDJHSXOM","created_at":"2026-05-25T02:01:21Z"},{"alias_kind":"pith_short_16","alias_value":"C67XSDJHSXOMZV4Z","created_at":"2026-05-25T02:01:21Z"},{"alias_kind":"pith_short_8","alias_value":"C67XSDJH","created_at":"2026-05-25T02:01:21Z"}],"graph_snapshots":[{"event_id":"sha256:dab93c22455c026eda5dd4fb10f3f63301b4bce6d03507c95d7106d7913d843f","target":"graph","created_at":"2026-05-25T02:01:21Z","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":4,"items":[{"attestation":"unclaimed","claim_id":"C1","kind":"strongest_claim","source":"verdict.strongest_claim","status":"machine_extracted","text":"Experimental results demonstrate that all strategies achieve sub-linear cumulative regret, with UCB-Bayes converging the fastest, followed by UCB-Tuned and UCB-V. Finally, UCB-V and UCB-Tuned dominate the Pareto Frontiers of accuracy-latency and accuracy-energy trade-offs."},{"attestation":"unclaimed","claim_id":"C2","kind":"weakest_assumption","source":"verdict.weakest_assumption","status":"machine_extracted","text":"The assumption that the multi-armed bandit reward distributions remain stationary across inference steps and that the chosen datasets and models adequately represent real edge-device conditions with varying input distributions and hardware constraints."},{"attestation":"unclaimed","claim_id":"C3","kind":"one_line_summary","source":"verdict.one_line_summary","status":"machine_extracted","text":"UCB-V and UCB-Tuned dominate accuracy-energy and accuracy-latency trade-offs while all tested UCB strategies achieve sub-linear regret in adaptive DNN early-exit experiments on CIFAR datasets."},{"attestation":"unclaimed","claim_id":"C4","kind":"headline","source":"verdict.pith_extraction.headline","status":"machine_extracted","text":"Multiple Upper Confidence Bound strategies achieve sub-linear regret in Adaptive Deep Neural Networks and improve accuracy-energy and accuracy-latency trade-offs."}],"snapshot_sha256":"3a540bb6895f01179888b83ca0fcd8977fc5e2e0f0cb31d6ed1bb22256a5ee1e"},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"integrity":{"available":true,"clean":true,"detectors_run":[{"findings_count":0,"name":"ai_meta_artifact","ran_at":"2026-05-21T07:36:22.860224Z","status":"completed","version":"1.0.0"},{"findings_count":0,"name":"doi_compliance","ran_at":"2026-05-19T22:23:14.663615Z","status":"completed","version":"1.0.0"}],"endpoint":"/pith/2604.24810/integrity.json","findings":[],"snapshot_sha256":"a119f36626d20b77ea381903fa106606cdb816b71db33090b17509c776b9f701","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Edge computing environments impose strict constraints on energy consumption and latency, making the deployment of deep neural networks a significant challenge. Therefore, smart and adaptive inference strategies that dynamically balance computational cost or latency with predictive accuracy are critical in edge computing scenarios. In this work, we build on Adaptive Deep Neural Networks (ADNNs) that employ the Multi-Armed Bandit (MAB) framework. Current literature leverages the first version of the Upper Confidence Bound (UCB1) strategy to dynamically select the optimal confidence threshold, en","authors_text":"Grigorios Papanikolaou, Ioannis Kontopoulos, Konstantinos Tserpes","cross_cats":["cs.AI"],"headline":"Multiple Upper Confidence Bound strategies achieve sub-linear regret in Adaptive Deep Neural Networks and improve accuracy-energy and accuracy-latency trade-offs.","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-04-27T08:51:44Z","title":"A Comparative Analysis on the Performance of Upper Confidence Bound Algorithms in Adaptive Deep Neural Networks"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2604.24810","kind":"arxiv","version":3},"verdict":{"created_at":"2026-05-08T04:20:33.911517Z","id":"96c191fb-8275-4f23-b425-b889be43d580","model_set":{"reader":"grok-4.3"},"one_line_summary":"UCB-V and UCB-Tuned dominate accuracy-energy and accuracy-latency trade-offs while all tested UCB strategies achieve sub-linear regret in adaptive DNN early-exit experiments on CIFAR datasets.","pipeline_version":"pith-pipeline@v0.9.0","pith_extraction_headline":"Multiple Upper Confidence Bound strategies achieve sub-linear regret in Adaptive Deep Neural Networks and improve accuracy-energy and accuracy-latency trade-offs.","strongest_claim":"Experimental results demonstrate that all strategies achieve sub-linear cumulative regret, with UCB-Bayes converging the fastest, followed by UCB-Tuned and UCB-V. Finally, UCB-V and UCB-Tuned dominate the Pareto Frontiers of accuracy-latency and accuracy-energy trade-offs.","weakest_assumption":"The assumption that the multi-armed bandit reward distributions remain stationary across inference steps and that the chosen datasets and models adequately represent real edge-device conditions with varying input distributions and hardware constraints."}},"verdict_id":"96c191fb-8275-4f23-b425-b889be43d580"}}],"author_attestations":[],"timestamp_anchors":[],"storage_attestations":[],"citation_signatures":[],"replication_records":[],"corrections":[],"mirror_hints":[],"record_created":{"event_id":"sha256:8b2cb1cb2fe3d470e55f74f8d709035c4dc4ea3beee937067cd799382200cda5","target":"record","created_at":"2026-05-25T02:01:21Z","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":"d3ef120d0fa5b8e02c510662b287655029ea01cca648c04094e654637d5a3f73","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-04-27T08:51:44Z","title_canon_sha256":"8298ff77409caa6bd2038e49aadafb2e4053ae1733440fa7aee06fd2fd961f44"},"schema_version":"1.0","source":{"id":"2604.24810","kind":"arxiv","version":3}},"canonical_sha256":"17bf790d2795dcccd7997bee59ff86449096c9bf46c34885e3d899c74d122dcd","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"17bf790d2795dcccd7997bee59ff86449096c9bf46c34885e3d899c74d122dcd","first_computed_at":"2026-05-25T02:01:21.158039Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-05-25T02:01:21.158039Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"aRt4MkFmn7xz8YraLho740nyU15LasqTv2PUdGRXa7BGWPYkNMbAIJ9Om9WSQC0Uh8BzOBBVOpek40nnvAh8Dg==","signature_status":"signed_v1","signed_at":"2026-05-25T02:01:21.158845Z","signed_message":"canonical_sha256_bytes"},"source_id":"2604.24810","source_kind":"arxiv","source_version":3}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:8b2cb1cb2fe3d470e55f74f8d709035c4dc4ea3beee937067cd799382200cda5","sha256:dab93c22455c026eda5dd4fb10f3f63301b4bce6d03507c95d7106d7913d843f"],"state_sha256":"3d01c2fe8cb8ae68643964544c21b73722592fe04c8c28358fd6e453c0537a2f"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"K38LACm0IjAsjtUjmyhFuaH5Bww4M+xo0NYGJuqc0EHXHTZFXa0gLSWwzmKU86EEDCeU2nlAmygeyhq82BDCDA==","signed_message":"bundle_sha256_bytes","signed_at":"2026-05-27T20:15:50.811550Z","bundle_sha256":"ca4a1a0f171d2f07a842bf640db2b3273791290746038db087fb1e13eb992c68"}}