{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:5RCVWDSAVRVJKT5BQ4UT6XP5EB","short_pith_number":"pith:5RCVWDSA","schema_version":"1.0","canonical_sha256":"ec455b0e40ac6a954fa187293f5dfd2074dd057674faf28622d96e066b7d790c","source":{"kind":"arxiv","id":"2502.02853","version":5},"attestation_state":"computed","paper":{"title":"Rethinking Latent Redundancy in Behavior Cloning: An Information Bottleneck Approach for Robot Manipulation","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.RO","authors_text":"Badong Chen, Donglin Wang, Pengxiang Ding, Shuanghao Bai, Wanqi Zhou, Wei Zhao","submitted_at":"2025-02-05T03:13:04Z","abstract_excerpt":"Behavior Cloning (BC) is a widely adopted visual imitation learning method in robot manipulation. Current BC approaches often enhance generalization by leveraging large datasets and incorporating additional visual and textual modalities to capture more diverse information. However, these methods overlook whether the learned representations contain redundant information and lack a solid theoretical foundation to guide the learning process. To address these limitations, we adopt an information-theoretic perspective and introduce mutual information to quantify and mitigate redundancy in latent re"},"verification_status":{"content_addressed":true,"pith_receipt":true,"author_attested":false,"weak_author_claims":0,"strong_author_claims":0,"externally_anchored":false,"storage_verified":false,"citation_signatures":0,"replication_records":0,"graph_snapshot":true,"references_resolved":false,"formal_links_present":false},"canonical_record":{"source":{"id":"2502.02853","kind":"arxiv","version":5},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.RO","submitted_at":"2025-02-05T03:13:04Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"a6666b73b950fbd0dbb09f77f1b347dd450ec76803d1d9f5e7b99badbea3bfd5","abstract_canon_sha256":"da2aeea58d2d328bfe95ed6f7e1ae1795e879f5a5d2b97bc7812aa8bf871fd5f"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:02:17.924385Z","signature_b64":"B3H6KRsar00X9wwlGffB4Ybz5tHrMjlvO3pQyaZuBc3uRrM1KmSlXO3ZeEsAUyIF15p1ZfUHtqHU1drysfvPDA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"ec455b0e40ac6a954fa187293f5dfd2074dd057674faf28622d96e066b7d790c","last_reissued_at":"2026-07-05T11:02:17.923904Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:02:17.923904Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Rethinking Latent Redundancy in Behavior Cloning: An Information Bottleneck Approach for Robot Manipulation","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.RO","authors_text":"Badong Chen, Donglin Wang, Pengxiang Ding, Shuanghao Bai, Wanqi Zhou, Wei Zhao","submitted_at":"2025-02-05T03:13:04Z","abstract_excerpt":"Behavior Cloning (BC) is a widely adopted visual imitation learning method in robot manipulation. Current BC approaches often enhance generalization by leveraging large datasets and incorporating additional visual and textual modalities to capture more diverse information. However, these methods overlook whether the learned representations contain redundant information and lack a solid theoretical foundation to guide the learning process. To address these limitations, we adopt an information-theoretic perspective and introduce mutual information to quantify and mitigate redundancy in latent re"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2502.02853","kind":"arxiv","version":5},"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/2502.02853/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"},"aliases":[{"alias_kind":"arxiv","alias_value":"2502.02853","created_at":"2026-07-05T11:02:17.923960+00:00"},{"alias_kind":"arxiv_version","alias_value":"2502.02853v5","created_at":"2026-07-05T11:02:17.923960+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2502.02853","created_at":"2026-07-05T11:02:17.923960+00:00"},{"alias_kind":"pith_short_12","alias_value":"5RCVWDSAVRVJ","created_at":"2026-07-05T11:02:17.923960+00:00"},{"alias_kind":"pith_short_16","alias_value":"5RCVWDSAVRVJKT5B","created_at":"2026-07-05T11:02:17.923960+00:00"},{"alias_kind":"pith_short_8","alias_value":"5RCVWDSA","created_at":"2026-07-05T11:02:17.923960+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":5,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.19998","citing_title":"Tri-Info: Generalizable, Interpretable Failure Prediction for VLA Models via Information Theory","ref_index":56,"is_internal_anchor":false},{"citing_arxiv_id":"2606.08653","citing_title":"FiberTune: Preserving Action-Fiber Visual Residuals in Vision-Language-Action Fine-Tuning","ref_index":26,"is_internal_anchor":false},{"citing_arxiv_id":"2606.29570","citing_title":"Hierarchical Policy Learning via Spectral Decomposition","ref_index":9,"is_internal_anchor":false},{"citing_arxiv_id":"2605.17486","citing_title":"DyGRO-VLA: Cross-Task Scaling of Vision-Language-Action Models via Dynamic Grouped Residual Optimization","ref_index":44,"is_internal_anchor":false},{"citing_arxiv_id":"2601.21926","citing_title":"Information Filtering via Variational Regularization for Robot Manipulation","ref_index":2,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/5RCVWDSAVRVJKT5BQ4UT6XP5EB","json":"https://pith.science/pith/5RCVWDSAVRVJKT5BQ4UT6XP5EB.json","graph_json":"https://pith.science/api/pith-number/5RCVWDSAVRVJKT5BQ4UT6XP5EB/graph.json","events_json":"https://pith.science/api/pith-number/5RCVWDSAVRVJKT5BQ4UT6XP5EB/events.json","paper":"https://pith.science/paper/5RCVWDSA"},"agent_actions":{"view_html":"https://pith.science/pith/5RCVWDSAVRVJKT5BQ4UT6XP5EB","download_json":"https://pith.science/pith/5RCVWDSAVRVJKT5BQ4UT6XP5EB.json","view_paper":"https://pith.science/paper/5RCVWDSA","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2502.02853&json=true","fetch_graph":"https://pith.science/api/pith-number/5RCVWDSAVRVJKT5BQ4UT6XP5EB/graph.json","fetch_events":"https://pith.science/api/pith-number/5RCVWDSAVRVJKT5BQ4UT6XP5EB/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/5RCVWDSAVRVJKT5BQ4UT6XP5EB/action/timestamp_anchor","attest_storage":"https://pith.science/pith/5RCVWDSAVRVJKT5BQ4UT6XP5EB/action/storage_attestation","attest_author":"https://pith.science/pith/5RCVWDSAVRVJKT5BQ4UT6XP5EB/action/author_attestation","sign_citation":"https://pith.science/pith/5RCVWDSAVRVJKT5BQ4UT6XP5EB/action/citation_signature","submit_replication":"https://pith.science/pith/5RCVWDSAVRVJKT5BQ4UT6XP5EB/action/replication_record"}},"created_at":"2026-07-05T11:02:17.923960+00:00","updated_at":"2026-07-05T11:02:17.923960+00:00"}