{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2018:P4YLH564SBGAU2A53JKAJWSVJU","short_pith_number":"pith:P4YLH564","schema_version":"1.0","canonical_sha256":"7f30b3f7dc904c0a681dda5404da554d1ccb647ab92ab30dd9315e06fa8338b5","source":{"kind":"arxiv","id":"1802.04942","version":5},"attestation_state":"computed","paper":{"title":"Isolating Sources of Disentanglement in Variational Autoencoders","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","stat.ML"],"primary_cat":"cs.LG","authors_text":"David Duvenaud, Ricky T. Q. Chen, Roger Grosse, Xuechen Li","submitted_at":"2018-02-14T03:48:06Z","abstract_excerpt":"We decompose the evidence lower bound to show the existence of a term measuring the total correlation between latent variables. We use this to motivate our $\\beta$-TCVAE (Total Correlation Variational Autoencoder), a refinement of the state-of-the-art $\\beta$-VAE objective for learning disentangled representations, requiring no additional hyperparameters during training. We further propose a principled classifier-free measure of disentanglement called the mutual information gap (MIG). We perform extensive quantitative and qualitative experiments, in both restricted and non-restricted settings,"},"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":"1802.04942","kind":"arxiv","version":5},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2018-02-14T03:48:06Z","cross_cats_sorted":["cs.AI","stat.ML"],"title_canon_sha256":"f7b416413e5c1e4f9911df24ef436f461cd0f2a85187f3dcec3008c937f2e399","abstract_canon_sha256":"107228fc8f13249f949e86ea8436d5dc33d9e16006280d9aa8a0625d089129f0"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-17T23:48:02.053234Z","signature_b64":"qGIvtu5g7ZsgNkxpDvepTa+QeAEsXguRA9ekTT4kAEsbHROHP4O7SO9vWvulIolfSPnTG1SfZQxxE1lsv1TvAQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"7f30b3f7dc904c0a681dda5404da554d1ccb647ab92ab30dd9315e06fa8338b5","last_reissued_at":"2026-05-17T23:48:02.052576Z","signature_status":"signed_v1","first_computed_at":"2026-05-17T23:48:02.052576Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Isolating Sources of Disentanglement in Variational Autoencoders","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","stat.ML"],"primary_cat":"cs.LG","authors_text":"David Duvenaud, Ricky T. Q. Chen, Roger Grosse, Xuechen Li","submitted_at":"2018-02-14T03:48:06Z","abstract_excerpt":"We decompose the evidence lower bound to show the existence of a term measuring the total correlation between latent variables. We use this to motivate our $\\beta$-TCVAE (Total Correlation Variational Autoencoder), a refinement of the state-of-the-art $\\beta$-VAE objective for learning disentangled representations, requiring no additional hyperparameters during training. We further propose a principled classifier-free measure of disentanglement called the mutual information gap (MIG). We perform extensive quantitative and qualitative experiments, in both restricted and non-restricted settings,"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"1802.04942","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":""},"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":"1802.04942","created_at":"2026-05-17T23:48:02.052683+00:00"},{"alias_kind":"arxiv_version","alias_value":"1802.04942v5","created_at":"2026-05-17T23:48:02.052683+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1802.04942","created_at":"2026-05-17T23:48:02.052683+00:00"},{"alias_kind":"pith_short_12","alias_value":"P4YLH564SBGA","created_at":"2026-05-18T12:32:43.782077+00:00"},{"alias_kind":"pith_short_16","alias_value":"P4YLH564SBGAU2A5","created_at":"2026-05-18T12:32:43.782077+00:00"},{"alias_kind":"pith_short_8","alias_value":"P4YLH564","created_at":"2026-05-18T12:32:43.782077+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2605.20732","citing_title":"Deep Attention Reweighting: Post-Hoc Attention-Based Feature Aggregation in CNNs for Disentangling Core and Spurious Features under Spurious Correlations","ref_index":8,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/P4YLH564SBGAU2A53JKAJWSVJU","json":"https://pith.science/pith/P4YLH564SBGAU2A53JKAJWSVJU.json","graph_json":"https://pith.science/api/pith-number/P4YLH564SBGAU2A53JKAJWSVJU/graph.json","events_json":"https://pith.science/api/pith-number/P4YLH564SBGAU2A53JKAJWSVJU/events.json","paper":"https://pith.science/paper/P4YLH564"},"agent_actions":{"view_html":"https://pith.science/pith/P4YLH564SBGAU2A53JKAJWSVJU","download_json":"https://pith.science/pith/P4YLH564SBGAU2A53JKAJWSVJU.json","view_paper":"https://pith.science/paper/P4YLH564","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=1802.04942&json=true","fetch_graph":"https://pith.science/api/pith-number/P4YLH564SBGAU2A53JKAJWSVJU/graph.json","fetch_events":"https://pith.science/api/pith-number/P4YLH564SBGAU2A53JKAJWSVJU/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/P4YLH564SBGAU2A53JKAJWSVJU/action/timestamp_anchor","attest_storage":"https://pith.science/pith/P4YLH564SBGAU2A53JKAJWSVJU/action/storage_attestation","attest_author":"https://pith.science/pith/P4YLH564SBGAU2A53JKAJWSVJU/action/author_attestation","sign_citation":"https://pith.science/pith/P4YLH564SBGAU2A53JKAJWSVJU/action/citation_signature","submit_replication":"https://pith.science/pith/P4YLH564SBGAU2A53JKAJWSVJU/action/replication_record"}},"created_at":"2026-05-17T23:48:02.052683+00:00","updated_at":"2026-05-17T23:48:02.052683+00:00"}