{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2026:PNJWVCVUQVC23U7427HTLMJI6U","short_pith_number":"pith:PNJWVCVU","canonical_record":{"source":{"id":"2605.14252","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-05-14T01:39:14Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"cac011e9bf399d63f81ec4ffa36ff93585db30e28a2df00d5889cfbbb6be4d92","abstract_canon_sha256":"3a618e38a7c34a476a855760bc7067fef1052a6bf211434949ced69d1493d0ea"},"schema_version":"1.0"},"canonical_sha256":"7b536a8ab48545add3fcd7cf35b128f50be79be4936b1df295b55b9bffb3333e","source":{"kind":"arxiv","id":"2605.14252","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2605.14252","created_at":"2026-05-17T23:39:10Z"},{"alias_kind":"arxiv_version","alias_value":"2605.14252v1","created_at":"2026-05-17T23:39:10Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.14252","created_at":"2026-05-17T23:39:10Z"},{"alias_kind":"pith_short_12","alias_value":"PNJWVCVUQVC2","created_at":"2026-05-18T12:33:37Z"},{"alias_kind":"pith_short_16","alias_value":"PNJWVCVUQVC23U74","created_at":"2026-05-18T12:33:37Z"},{"alias_kind":"pith_short_8","alias_value":"PNJWVCVU","created_at":"2026-05-18T12:33:37Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2026:PNJWVCVUQVC23U7427HTLMJI6U","target":"record","payload":{"canonical_record":{"source":{"id":"2605.14252","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-05-14T01:39:14Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"cac011e9bf399d63f81ec4ffa36ff93585db30e28a2df00d5889cfbbb6be4d92","abstract_canon_sha256":"3a618e38a7c34a476a855760bc7067fef1052a6bf211434949ced69d1493d0ea"},"schema_version":"1.0"},"canonical_sha256":"7b536a8ab48545add3fcd7cf35b128f50be79be4936b1df295b55b9bffb3333e","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-17T23:39:10.556978Z","signature_b64":"lMb/utWuTix66slEi+9EUz4l1/uWhYHgEuRN7AeHH9x2aImrPk7tbiXat3JGbFoP15TkktAr6n4yvOofpZc/CA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"7b536a8ab48545add3fcd7cf35b128f50be79be4936b1df295b55b9bffb3333e","last_reissued_at":"2026-05-17T23:39:10.556516Z","signature_status":"signed_v1","first_computed_at":"2026-05-17T23:39:10.556516Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2605.14252","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":"2026-05-17T23:39:10Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"wfzWvZZe27libkERlFd0Ml31GcIwEOHrBjbpnJ614RltSzE7E4kz6YS7C0BqXnBquBxQMQRQrn/OIC01B4C6Cw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-05-23T04:19:53.964398Z"},"content_sha256":"bb6f73d39cb8e9dd49fb8a49ab6188558895aa0015a89290b4377982d3f03bab","schema_version":"1.0","event_id":"sha256:bb6f73d39cb8e9dd49fb8a49ab6188558895aa0015a89290b4377982d3f03bab"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2026:PNJWVCVUQVC23U7427HTLMJI6U","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Not All Timesteps Matter Equally: Selective Alignment Knowledge Distillation for Spiking Neural Networks","license":"http://creativecommons.org/licenses/by/4.0/","headline":"Spiking neural networks gain accuracy when distillation corrects only erroneous timesteps instead of aligning every one uniformly.","cross_cats":["cs.AI"],"primary_cat":"cs.LG","authors_text":"Benjamin Smith, Guowei Zhang, Kai Sun, Levin Kuhlmann, Nanxu Gong, Peibo Duan, Yongsheng Huang","submitted_at":"2026-05-14T01:39:14Z","abstract_excerpt":"Spiking neural networks (SNNs), which are brain-inspired and spike-driven, achieve high energy efficiency. However, a performance gap between SNNs and artificial neural networks (ANNs) still remains. Knowledge distillation (KD) is commonly adopted to improve SNN performance, but existing methods typically enforce uniform alignment across all timesteps, either from a teacher network or through inter-temporal self-distillation, implicitly assuming that per-timestep predictions should be treated equally. In practice, SNN predictions vary and evolve over time, and intermediate timesteps need not a"},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"Extensive experiments on static image and neuromorphic event-based datasets demonstrate consistent improvements over existing distillation methods.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"That erroneous timesteps can be reliably identified from the student's own outputs without introducing new hyperparameters that themselves require extensive tuning or that the reweighting scheme based on confidence and similarity is robust across datasets.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"SeAl-KD improves SNN accuracy by equalizing competing logits at erroneous timesteps and reweighting temporal alignment using confidence and inter-timestep similarity.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"Spiking neural networks gain accuracy when distillation corrects only erroneous timesteps instead of aligning every one uniformly.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"b52f595919409167ec632b040610c642404ae9f3159016f87df7fe0462d7ebaa"},"source":{"id":"2605.14252","kind":"arxiv","version":1},"verdict":{"id":"f2335ebe-5253-48c3-9dc0-60ad1a5453cf","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-15T02:41:22.063761Z","strongest_claim":"Extensive experiments on static image and neuromorphic event-based datasets demonstrate consistent improvements over existing distillation methods.","one_line_summary":"SeAl-KD improves SNN accuracy by equalizing competing logits at erroneous timesteps and reweighting temporal alignment using confidence and inter-timestep similarity.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"That erroneous timesteps can be reliably identified from the student's own outputs without introducing new hyperparameters that themselves require extensive tuning or that the reweighting scheme based on confidence and similarity is robust across datasets.","pith_extraction_headline":"Spiking neural networks gain accuracy when distillation corrects only erroneous timesteps instead of aligning every one uniformly."},"references":{"count":30,"sample":[{"doi":"","year":2018,"title":"Long short-term memory and learning-to-learn in net- works of spiking neurons.Advances in neural information processing systems, 31,","work_id":"eea36dbe-89d9-4295-9f60-6770ac91463b","ref_index":1,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2009,"title":"Imagenet: A large-scale hierarchical image database","work_id":"681c738f-6bc1-4df9-a855-53a601d517a9","ref_index":2,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2022,"title":"Temporal efficient training of spiking neural network via gradient re-weighting","work_id":"9edf39ac-b8ed-46fc-8d9d-0df24480f028","ref_index":3,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2022,"title":"Temporal effective batch normalization in spiking neural networks.Advances in Neural Information Processing Systems, 35:34377– 34390,","work_id":"65401ea0-bdc3-40d7-8b06-2f363cfa0995","ref_index":4,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2021,"title":"In- corporating learnable membrane time constant to enhance learning of spiking neural networks","work_id":"1b1a2299-8d72-498a-9468-1a1bfb3e2182","ref_index":5,"cited_arxiv_id":"","is_internal_anchor":false}],"resolved_work":30,"snapshot_sha256":"3e6a89dbbe29c80956ce634ed4fc2aea3c2c723952c30054cf7b20289f4cee26","internal_anchors":0},"formal_canon":{"evidence_count":2,"snapshot_sha256":"ac773b479f5d2d849d89c6e733705b5a941baacde8671722493f4239c1d65ae7"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"},"verdict_id":"f2335ebe-5253-48c3-9dc0-60ad1a5453cf"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-05-17T23:39:10Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"D8vSVr1Emtuv2/N9o+oMb3vMNqClVmnPhP3bPfKSwrZlnCIGyDTHONrQRodo7rghqS8N3wTPDMrOug82f6R2BA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-05-23T04:19:53.965089Z"},"content_sha256":"5afd5435f9d772ca3c72e530c8866d548df511473832fd16a4bbbe89b99b8568","schema_version":"1.0","event_id":"sha256:5afd5435f9d772ca3c72e530c8866d548df511473832fd16a4bbbe89b99b8568"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/PNJWVCVUQVC23U7427HTLMJI6U/bundle.json","state_url":"https://pith.science/pith/PNJWVCVUQVC23U7427HTLMJI6U/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/PNJWVCVUQVC23U7427HTLMJI6U/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":"2026-05-23T04:19:53Z","links":{"resolver":"https://pith.science/pith/PNJWVCVUQVC23U7427HTLMJI6U","bundle":"https://pith.science/pith/PNJWVCVUQVC23U7427HTLMJI6U/bundle.json","state":"https://pith.science/pith/PNJWVCVUQVC23U7427HTLMJI6U/state.json","well_known_bundle":"https://pith.science/.well-known/pith/PNJWVCVUQVC23U7427HTLMJI6U/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:PNJWVCVUQVC23U7427HTLMJI6U","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":"3a618e38a7c34a476a855760bc7067fef1052a6bf211434949ced69d1493d0ea","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-05-14T01:39:14Z","title_canon_sha256":"cac011e9bf399d63f81ec4ffa36ff93585db30e28a2df00d5889cfbbb6be4d92"},"schema_version":"1.0","source":{"id":"2605.14252","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2605.14252","created_at":"2026-05-17T23:39:10Z"},{"alias_kind":"arxiv_version","alias_value":"2605.14252v1","created_at":"2026-05-17T23:39:10Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.14252","created_at":"2026-05-17T23:39:10Z"},{"alias_kind":"pith_short_12","alias_value":"PNJWVCVUQVC2","created_at":"2026-05-18T12:33:37Z"},{"alias_kind":"pith_short_16","alias_value":"PNJWVCVUQVC23U74","created_at":"2026-05-18T12:33:37Z"},{"alias_kind":"pith_short_8","alias_value":"PNJWVCVU","created_at":"2026-05-18T12:33:37Z"}],"graph_snapshots":[{"event_id":"sha256:5afd5435f9d772ca3c72e530c8866d548df511473832fd16a4bbbe89b99b8568","target":"graph","created_at":"2026-05-17T23:39:10Z","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":"Extensive experiments on static image and neuromorphic event-based datasets demonstrate consistent improvements over existing distillation methods."},{"attestation":"unclaimed","claim_id":"C2","kind":"weakest_assumption","source":"verdict.weakest_assumption","status":"machine_extracted","text":"That erroneous timesteps can be reliably identified from the student's own outputs without introducing new hyperparameters that themselves require extensive tuning or that the reweighting scheme based on confidence and similarity is robust across datasets."},{"attestation":"unclaimed","claim_id":"C3","kind":"one_line_summary","source":"verdict.one_line_summary","status":"machine_extracted","text":"SeAl-KD improves SNN accuracy by equalizing competing logits at erroneous timesteps and reweighting temporal alignment using confidence and inter-timestep similarity."},{"attestation":"unclaimed","claim_id":"C4","kind":"headline","source":"verdict.pith_extraction.headline","status":"machine_extracted","text":"Spiking neural networks gain accuracy when distillation corrects only erroneous timesteps instead of aligning every one uniformly."}],"snapshot_sha256":"b52f595919409167ec632b040610c642404ae9f3159016f87df7fe0462d7ebaa"},"formal_canon":{"evidence_count":2,"snapshot_sha256":"ac773b479f5d2d849d89c6e733705b5a941baacde8671722493f4239c1d65ae7"},"paper":{"abstract_excerpt":"Spiking neural networks (SNNs), which are brain-inspired and spike-driven, achieve high energy efficiency. However, a performance gap between SNNs and artificial neural networks (ANNs) still remains. Knowledge distillation (KD) is commonly adopted to improve SNN performance, but existing methods typically enforce uniform alignment across all timesteps, either from a teacher network or through inter-temporal self-distillation, implicitly assuming that per-timestep predictions should be treated equally. In practice, SNN predictions vary and evolve over time, and intermediate timesteps need not a","authors_text":"Benjamin Smith, Guowei Zhang, Kai Sun, Levin Kuhlmann, Nanxu Gong, Peibo Duan, Yongsheng Huang","cross_cats":["cs.AI"],"headline":"Spiking neural networks gain accuracy when distillation corrects only erroneous timesteps instead of aligning every one uniformly.","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-05-14T01:39:14Z","title":"Not All Timesteps Matter Equally: Selective Alignment Knowledge Distillation for Spiking Neural Networks"},"references":{"count":30,"internal_anchors":0,"resolved_work":30,"sample":[{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":1,"title":"Long short-term memory and learning-to-learn in net- works of spiking neurons.Advances in neural information processing systems, 31,","work_id":"eea36dbe-89d9-4295-9f60-6770ac91463b","year":2018},{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":2,"title":"Imagenet: A large-scale hierarchical image database","work_id":"681c738f-6bc1-4df9-a855-53a601d517a9","year":2009},{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":3,"title":"Temporal efficient training of spiking neural network via gradient re-weighting","work_id":"9edf39ac-b8ed-46fc-8d9d-0df24480f028","year":2022},{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":4,"title":"Temporal effective batch normalization in spiking neural networks.Advances in Neural Information Processing Systems, 35:34377– 34390,","work_id":"65401ea0-bdc3-40d7-8b06-2f363cfa0995","year":2022},{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":5,"title":"In- corporating learnable membrane time constant to enhance learning of spiking neural networks","work_id":"1b1a2299-8d72-498a-9468-1a1bfb3e2182","year":2021}],"snapshot_sha256":"3e6a89dbbe29c80956ce634ed4fc2aea3c2c723952c30054cf7b20289f4cee26"},"source":{"id":"2605.14252","kind":"arxiv","version":1},"verdict":{"created_at":"2026-05-15T02:41:22.063761Z","id":"f2335ebe-5253-48c3-9dc0-60ad1a5453cf","model_set":{"reader":"grok-4.3"},"one_line_summary":"SeAl-KD improves SNN accuracy by equalizing competing logits at erroneous timesteps and reweighting temporal alignment using confidence and inter-timestep similarity.","pipeline_version":"pith-pipeline@v0.9.0","pith_extraction_headline":"Spiking neural networks gain accuracy when distillation corrects only erroneous timesteps instead of aligning every one uniformly.","strongest_claim":"Extensive experiments on static image and neuromorphic event-based datasets demonstrate consistent improvements over existing distillation methods.","weakest_assumption":"That erroneous timesteps can be reliably identified from the student's own outputs without introducing new hyperparameters that themselves require extensive tuning or that the reweighting scheme based on confidence and similarity is robust across datasets."}},"verdict_id":"f2335ebe-5253-48c3-9dc0-60ad1a5453cf"}}],"author_attestations":[],"timestamp_anchors":[],"storage_attestations":[],"citation_signatures":[],"replication_records":[],"corrections":[],"mirror_hints":[],"record_created":{"event_id":"sha256:bb6f73d39cb8e9dd49fb8a49ab6188558895aa0015a89290b4377982d3f03bab","target":"record","created_at":"2026-05-17T23:39:10Z","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":"3a618e38a7c34a476a855760bc7067fef1052a6bf211434949ced69d1493d0ea","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-05-14T01:39:14Z","title_canon_sha256":"cac011e9bf399d63f81ec4ffa36ff93585db30e28a2df00d5889cfbbb6be4d92"},"schema_version":"1.0","source":{"id":"2605.14252","kind":"arxiv","version":1}},"canonical_sha256":"7b536a8ab48545add3fcd7cf35b128f50be79be4936b1df295b55b9bffb3333e","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"7b536a8ab48545add3fcd7cf35b128f50be79be4936b1df295b55b9bffb3333e","first_computed_at":"2026-05-17T23:39:10.556516Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-05-17T23:39:10.556516Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"lMb/utWuTix66slEi+9EUz4l1/uWhYHgEuRN7AeHH9x2aImrPk7tbiXat3JGbFoP15TkktAr6n4yvOofpZc/CA==","signature_status":"signed_v1","signed_at":"2026-05-17T23:39:10.556978Z","signed_message":"canonical_sha256_bytes"},"source_id":"2605.14252","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:bb6f73d39cb8e9dd49fb8a49ab6188558895aa0015a89290b4377982d3f03bab","sha256:5afd5435f9d772ca3c72e530c8866d548df511473832fd16a4bbbe89b99b8568"],"state_sha256":"923632ea5d0a3040f1185909bd46920b72cf8fb0354a646c6d06e2954a3cafa7"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"3Qc5uBhQ4xqw8K9/XTvNtBr5gFagoqLv1Xsd/+lXw8UNFHzNcu1qB8QbDMWwhpMeIsD8vGwERIvL7jW7FgxECA==","signed_message":"bundle_sha256_bytes","signed_at":"2026-05-23T04:19:53.968481Z","bundle_sha256":"1f81efd13dc9f9cb0956ac61e6aa8e060e90a1709c1483d0eed427c227d845eb"}}