{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2022:7YHRQHFB2L6L3QZL5AGWARVLT3","short_pith_number":"pith:7YHRQHFB","schema_version":"1.0","canonical_sha256":"fe0f181ca1d2fcbdc32be80d6046ab9ee83059ecf209b0cad247d8d7f98dd69d","source":{"kind":"arxiv","id":"2208.04202","version":2},"attestation_state":"computed","paper":{"title":"Analog Bits: Generating Discrete Data using Diffusion Models with Self-Conditioning","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.CL","cs.LG"],"primary_cat":"cs.CV","authors_text":"Geoffrey Hinton, Ruixiang Zhang, Ting Chen","submitted_at":"2022-08-08T15:08:40Z","abstract_excerpt":"We present Bit Diffusion: a simple and generic approach for generating discrete data with continuous state and continuous time diffusion models. The main idea behind our approach is to first represent the discrete data as binary bits, and then train a continuous diffusion model to model these bits as real numbers which we call analog bits. To generate samples, the model first generates the analog bits, which are then thresholded to obtain the bits that represent the discrete variables. We further propose two simple techniques, namely Self-Conditioning and Asymmetric Time Intervals, which lead "},"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":"2208.04202","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2022-08-08T15:08:40Z","cross_cats_sorted":["cs.AI","cs.CL","cs.LG"],"title_canon_sha256":"85910f3dca9d038fc6f298ae710bbce2f2a5d89a7c842948141943b529c61252","abstract_canon_sha256":"71cc99dc13f9cf759f99cc1c98c272530622127ef23eb9beb2fc7844dd143906"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T05:46:51.744569Z","signature_b64":"kbS0K8ZaqVAINv1DR3XJqaWdUF1wKrq7qBYxqC1EcegXD8wo0juJmOTzIlsr5Akz6au9a7fKd8NV/9wkCSqmCA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"fe0f181ca1d2fcbdc32be80d6046ab9ee83059ecf209b0cad247d8d7f98dd69d","last_reissued_at":"2026-07-05T05:46:51.744067Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T05:46:51.744067Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Analog Bits: Generating Discrete Data using Diffusion Models with Self-Conditioning","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.CL","cs.LG"],"primary_cat":"cs.CV","authors_text":"Geoffrey Hinton, Ruixiang Zhang, Ting Chen","submitted_at":"2022-08-08T15:08:40Z","abstract_excerpt":"We present Bit Diffusion: a simple and generic approach for generating discrete data with continuous state and continuous time diffusion models. The main idea behind our approach is to first represent the discrete data as binary bits, and then train a continuous diffusion model to model these bits as real numbers which we call analog bits. To generate samples, the model first generates the analog bits, which are then thresholded to obtain the bits that represent the discrete variables. We further propose two simple techniques, namely Self-Conditioning and Asymmetric Time Intervals, which lead "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2208.04202","kind":"arxiv","version":2},"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/2208.04202/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":"2208.04202","created_at":"2026-07-05T05:46:51.744124+00:00"},{"alias_kind":"arxiv_version","alias_value":"2208.04202v2","created_at":"2026-07-05T05:46:51.744124+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2208.04202","created_at":"2026-07-05T05:46:51.744124+00:00"},{"alias_kind":"pith_short_12","alias_value":"7YHRQHFB2L6L","created_at":"2026-07-05T05:46:51.744124+00:00"},{"alias_kind":"pith_short_16","alias_value":"7YHRQHFB2L6L3QZL","created_at":"2026-07-05T05:46:51.744124+00:00"},{"alias_kind":"pith_short_8","alias_value":"7YHRQHFB","created_at":"2026-07-05T05:46:51.744124+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":28,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2607.06930","citing_title":"Imputation Meets Clustering: Exploiting Latent Subgroup Structure for Missing Data Recovery","ref_index":5,"is_internal_anchor":true},{"citing_arxiv_id":"2606.22702","citing_title":"Modular Diffusion Models for Structured Visual Recognition","ref_index":3,"is_internal_anchor":false},{"citing_arxiv_id":"2606.21061","citing_title":"Neural Architecture Distributions: A New Paradigm for Stochastic Segmentation","ref_index":34,"is_internal_anchor":false},{"citing_arxiv_id":"2606.20404","citing_title":"FlowBender: Feedback-Aware Training for Self-Correcting Conditional Flows","ref_index":9,"is_internal_anchor":false},{"citing_arxiv_id":"2607.01775","citing_title":"Set Diffusion: Interpolating Token Orderings Between Autoregression and Diffusion for Fast and Flexible Decoding","ref_index":48,"is_internal_anchor":false},{"citing_arxiv_id":"2606.08150","citing_title":"Property-Informed Diffusion-Based Text-to-Microstructure Generation","ref_index":7,"is_internal_anchor":false},{"citing_arxiv_id":"2607.00588","citing_title":"Low Perplexity is Repetition: A One-Dimensional Self-Conditioning Attractor in Continuous Diffusion LMs","ref_index":2,"is_internal_anchor":false},{"citing_arxiv_id":"2607.00714","citing_title":"Self-conditioned Flow Map Language Models via Fixed-point Flows","ref_index":3,"is_internal_anchor":false},{"citing_arxiv_id":"2606.02133","citing_title":"Variational Learning for Insertion-based Generation","ref_index":43,"is_internal_anchor":false},{"citing_arxiv_id":"2606.27617","citing_title":"Masked Language Flow Models","ref_index":9,"is_internal_anchor":false},{"citing_arxiv_id":"2606.29150","citing_title":"Flow Reasoning Models: Scaling Reasoning Through Iterative Self-Refinement","ref_index":5,"is_internal_anchor":false},{"citing_arxiv_id":"2605.29233","citing_title":"BlockBatch: Multi-Scale Consensus Decoding for Efficient Diffusion Language Model Inference","ref_index":2,"is_internal_anchor":false},{"citing_arxiv_id":"2602.04883","citing_title":"Protein Autoregressive Modeling via Multiscale Structure Generation","ref_index":9,"is_internal_anchor":false},{"citing_arxiv_id":"2602.16813","citing_title":"Flow Map Language Models: One-step Language Modeling via Continuous Denoising","ref_index":38,"is_internal_anchor":false},{"citing_arxiv_id":"2210.08933","citing_title":"DiffuSeq: Sequence to Sequence Text Generation with Diffusion Models","ref_index":1,"is_internal_anchor":false},{"citing_arxiv_id":"2605.19726","citing_title":"Efficient Long-Context Modeling in Diffusion Language Models via Block Approximate Sparse Attention","ref_index":4,"is_internal_anchor":false},{"citing_arxiv_id":"2211.15089","citing_title":"Continuous diffusion for categorical data","ref_index":11,"is_internal_anchor":false},{"citing_arxiv_id":"2505.16933","citing_title":"LLaDA-V: Large Language Diffusion Models with Visual Instruction Tuning","ref_index":89,"is_internal_anchor":false},{"citing_arxiv_id":"2309.16797","citing_title":"Promptbreeder: Self-Referential Self-Improvement Via Prompt Evolution","ref_index":75,"is_internal_anchor":false},{"citing_arxiv_id":"2602.16813","citing_title":"Flow Map Language Models: One-step Language Modeling via Continuous Denoising","ref_index":38,"is_internal_anchor":false},{"citing_arxiv_id":"2310.16834","citing_title":"Discrete Diffusion Modeling by Estimating the Ratios of the Data Distribution","ref_index":1,"is_internal_anchor":false},{"citing_arxiv_id":"2605.11577","citing_title":"BitLM: Unlocking Multi-Token Language Generation with Bitwise Continuous Diffusion","ref_index":6,"is_internal_anchor":false},{"citing_arxiv_id":"2604.26985","citing_title":"Simple Self-Conditioning Adaptation for Masked Diffusion Models","ref_index":16,"is_internal_anchor":false},{"citing_arxiv_id":"2604.18995","citing_title":"$R^2$-dLLM: Accelerating Diffusion Large Language Models via Spatio-Temporal Redundancy Reduction","ref_index":3,"is_internal_anchor":false},{"citing_arxiv_id":"2604.18966","citing_title":"Self-Improving Tabular Language Models via Iterative Reward-Guided Post-Training","ref_index":201,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/7YHRQHFB2L6L3QZL5AGWARVLT3","json":"https://pith.science/pith/7YHRQHFB2L6L3QZL5AGWARVLT3.json","graph_json":"https://pith.science/api/pith-number/7YHRQHFB2L6L3QZL5AGWARVLT3/graph.json","events_json":"https://pith.science/api/pith-number/7YHRQHFB2L6L3QZL5AGWARVLT3/events.json","paper":"https://pith.science/paper/7YHRQHFB"},"agent_actions":{"view_html":"https://pith.science/pith/7YHRQHFB2L6L3QZL5AGWARVLT3","download_json":"https://pith.science/pith/7YHRQHFB2L6L3QZL5AGWARVLT3.json","view_paper":"https://pith.science/paper/7YHRQHFB","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2208.04202&json=true","fetch_graph":"https://pith.science/api/pith-number/7YHRQHFB2L6L3QZL5AGWARVLT3/graph.json","fetch_events":"https://pith.science/api/pith-number/7YHRQHFB2L6L3QZL5AGWARVLT3/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/7YHRQHFB2L6L3QZL5AGWARVLT3/action/timestamp_anchor","attest_storage":"https://pith.science/pith/7YHRQHFB2L6L3QZL5AGWARVLT3/action/storage_attestation","attest_author":"https://pith.science/pith/7YHRQHFB2L6L3QZL5AGWARVLT3/action/author_attestation","sign_citation":"https://pith.science/pith/7YHRQHFB2L6L3QZL5AGWARVLT3/action/citation_signature","submit_replication":"https://pith.science/pith/7YHRQHFB2L6L3QZL5AGWARVLT3/action/replication_record"}},"created_at":"2026-07-05T05:46:51.744124+00:00","updated_at":"2026-07-05T05:46:51.744124+00:00"}