{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2019:VHZAM7LAWW6B34Q66H3CNORT55","short_pith_number":"pith:VHZAM7LA","schema_version":"1.0","canonical_sha256":"a9f2067d60b5bc1df21ef1f626ba33ef53ed96f1c47ff0a84e25b50dd282a893","source":{"kind":"arxiv","id":"1912.02164","version":4},"attestation_state":"computed","paper":{"title":"Plug and Play Language Models: A Simple Approach to Controlled Text Generation","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.LG"],"primary_cat":"cs.CL","authors_text":"Andrea Madotto, Eric Frank, Jane Hung, Janice Lan, Jason Yosinski, Piero Molino, Rosanne Liu, Sumanth Dathathri","submitted_at":"2019-12-04T18:32:15Z","abstract_excerpt":"Large transformer-based language models (LMs) trained on huge text corpora have shown unparalleled generation capabilities. However, controlling attributes of the generated language (e.g. switching topic or sentiment) is difficult without modifying the model architecture or fine-tuning on attribute-specific data and entailing the significant cost of retraining. We propose a simple alternative: the Plug and Play Language Model (PPLM) for controllable language generation, which combines a pretrained LM with one or more simple attribute classifiers that guide text generation without any further t"},"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":"1912.02164","kind":"arxiv","version":4},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2019-12-04T18:32:15Z","cross_cats_sorted":["cs.AI","cs.LG"],"title_canon_sha256":"de37b76705e7f5d1d4eb58b0a64bafa070a92237d186795ca974628408ed4c5b","abstract_canon_sha256":"6b5035137afca72b9ebb5bf15254673e7f9bedf8c1331b4dce0eefca7d0f33dd"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T00:45:16.454349Z","signature_b64":"cbM4l4b9A+prYot+hzI7rNW7JP2NJ3ns9KYyCl5OSpvzh/rZbYYg+/WossmZwaAjIHDHwehRZkniBhgCrePxCw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"a9f2067d60b5bc1df21ef1f626ba33ef53ed96f1c47ff0a84e25b50dd282a893","last_reissued_at":"2026-07-05T00:45:16.453863Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T00:45:16.453863Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Plug and Play Language Models: A Simple Approach to Controlled Text Generation","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.LG"],"primary_cat":"cs.CL","authors_text":"Andrea Madotto, Eric Frank, Jane Hung, Janice Lan, Jason Yosinski, Piero Molino, Rosanne Liu, Sumanth Dathathri","submitted_at":"2019-12-04T18:32:15Z","abstract_excerpt":"Large transformer-based language models (LMs) trained on huge text corpora have shown unparalleled generation capabilities. However, controlling attributes of the generated language (e.g. switching topic or sentiment) is difficult without modifying the model architecture or fine-tuning on attribute-specific data and entailing the significant cost of retraining. We propose a simple alternative: the Plug and Play Language Model (PPLM) for controllable language generation, which combines a pretrained LM with one or more simple attribute classifiers that guide text generation without any further t"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"1912.02164","kind":"arxiv","version":4},"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/1912.02164/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":"1912.02164","created_at":"2026-07-05T00:45:16.453919+00:00"},{"alias_kind":"arxiv_version","alias_value":"1912.02164v4","created_at":"2026-07-05T00:45:16.453919+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1912.02164","created_at":"2026-07-05T00:45:16.453919+00:00"},{"alias_kind":"pith_short_12","alias_value":"VHZAM7LAWW6B","created_at":"2026-07-05T00:45:16.453919+00:00"},{"alias_kind":"pith_short_16","alias_value":"VHZAM7LAWW6B34Q6","created_at":"2026-07-05T00:45:16.453919+00:00"},{"alias_kind":"pith_short_8","alias_value":"VHZAM7LA","created_at":"2026-07-05T00:45:16.453919+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":23,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2607.08071","citing_title":"COBART: Controlled, Optimized, Bidirectional and Auto-Regressive Transformer for Ad Headline Generation","ref_index":4,"is_internal_anchor":true},{"citing_arxiv_id":"2606.12234","citing_title":"On The Effectiveness-Fluency Trade-Off In LLM Conditioning: A Systematic Study","ref_index":36,"is_internal_anchor":false},{"citing_arxiv_id":"2607.00009","citing_title":"Controllable Narrative Rendering for Enhanced Assisted Writing","ref_index":13,"is_internal_anchor":false},{"citing_arxiv_id":"2606.07915","citing_title":"EditSR: Enhancing Neural Symbolic Regression via Edit-based Rectification","ref_index":58,"is_internal_anchor":false},{"citing_arxiv_id":"2605.25123","citing_title":"Inference-Time Alignment of Diffusion Models via Trust-Region Iterative Twisted Sequential Monte Carlo","ref_index":10,"is_internal_anchor":false},{"citing_arxiv_id":"2606.29201","citing_title":"Behavior Uncloning: Distilling Mode Redirection into Policy Weights without Inference-Time Steering","ref_index":7,"is_internal_anchor":false},{"citing_arxiv_id":"2605.25619","citing_title":"Analogies between Transformer Layers and Power Method","ref_index":5,"is_internal_anchor":false},{"citing_arxiv_id":"2605.27997","citing_title":"Where Does Toxicity Live? Mechanistic Localization and Targeted Suppression in Language Models","ref_index":8,"is_internal_anchor":false},{"citing_arxiv_id":"2605.23040","citing_title":"Steered Generation via Gradient-Based Optimization on Sparse Query Features","ref_index":10,"is_internal_anchor":false},{"citing_arxiv_id":"2605.17231","citing_title":"FishBack: Pullback Fisher Geometry for Optimal Activation Steering in Transformers","ref_index":34,"is_internal_anchor":false},{"citing_arxiv_id":"2605.19940","citing_title":"Robotics-Inspired Guardrails for Foundation Models in Socially Sensitive Domains","ref_index":10,"is_internal_anchor":false},{"citing_arxiv_id":"2106.01345","citing_title":"Decision Transformer: Reinforcement Learning via Sequence Modeling","ref_index":63,"is_internal_anchor":false},{"citing_arxiv_id":"2104.08773","citing_title":"Cross-Task Generalization via Natural Language Crowdsourcing Instructions","ref_index":21,"is_internal_anchor":false},{"citing_arxiv_id":"2602.10346","citing_title":"Geometry-Aware Decoding with Wasserstein-Regularized Truncation and Mass Penalties for Large Language Models","ref_index":7,"is_internal_anchor":false},{"citing_arxiv_id":"2605.13858","citing_title":"A Hormone-inspired Emotion Layer for Transformer language models (HELT)","ref_index":14,"is_internal_anchor":false},{"citing_arxiv_id":"2605.14004","citing_title":"Conditional Attribute Estimation with Autoregressive Sequence Models","ref_index":14,"is_internal_anchor":false},{"citing_arxiv_id":"2605.13101","citing_title":"Margin-calibrated Classifier Guidance for Property-driven Synthesis Planning","ref_index":101,"is_internal_anchor":false},{"citing_arxiv_id":"2605.12395","citing_title":"A Comparative Study of Controlled Text Generation Systems Using Level-Playing-Field Evaluation Principles","ref_index":13,"is_internal_anchor":false},{"citing_arxiv_id":"2605.12412","citing_title":"Stories in Space: In-Context Learning Trajectories in Conceptual Belief Space","ref_index":72,"is_internal_anchor":false},{"citing_arxiv_id":"2605.12345","citing_title":"Output Composability of QLoRA PEFT Modules for Plug-and-Play Attribute-Controlled Text Generation","ref_index":25,"is_internal_anchor":false},{"citing_arxiv_id":"2605.09995","citing_title":"Annotations Mitigate Post-Training Mode Collapse","ref_index":30,"is_internal_anchor":false},{"citing_arxiv_id":"2605.07631","citing_title":"Inference Time Causal Probing in LLMs","ref_index":5,"is_internal_anchor":false},{"citing_arxiv_id":"2605.07701","citing_title":"Guidance Is Not a Hyperparameter: Learning Dynamic Control in Diffusion Language Models","ref_index":15,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/VHZAM7LAWW6B34Q66H3CNORT55","json":"https://pith.science/pith/VHZAM7LAWW6B34Q66H3CNORT55.json","graph_json":"https://pith.science/api/pith-number/VHZAM7LAWW6B34Q66H3CNORT55/graph.json","events_json":"https://pith.science/api/pith-number/VHZAM7LAWW6B34Q66H3CNORT55/events.json","paper":"https://pith.science/paper/VHZAM7LA"},"agent_actions":{"view_html":"https://pith.science/pith/VHZAM7LAWW6B34Q66H3CNORT55","download_json":"https://pith.science/pith/VHZAM7LAWW6B34Q66H3CNORT55.json","view_paper":"https://pith.science/paper/VHZAM7LA","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=1912.02164&json=true","fetch_graph":"https://pith.science/api/pith-number/VHZAM7LAWW6B34Q66H3CNORT55/graph.json","fetch_events":"https://pith.science/api/pith-number/VHZAM7LAWW6B34Q66H3CNORT55/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/VHZAM7LAWW6B34Q66H3CNORT55/action/timestamp_anchor","attest_storage":"https://pith.science/pith/VHZAM7LAWW6B34Q66H3CNORT55/action/storage_attestation","attest_author":"https://pith.science/pith/VHZAM7LAWW6B34Q66H3CNORT55/action/author_attestation","sign_citation":"https://pith.science/pith/VHZAM7LAWW6B34Q66H3CNORT55/action/citation_signature","submit_replication":"https://pith.science/pith/VHZAM7LAWW6B34Q66H3CNORT55/action/replication_record"}},"created_at":"2026-07-05T00:45:16.453919+00:00","updated_at":"2026-07-05T00:45:16.453919+00:00"}