{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2021:WE3DIJH5AKOVI7ADOSQ4SBAJII","short_pith_number":"pith:WE3DIJH5","schema_version":"1.0","canonical_sha256":"b1363424fd029d547c0374a1c904094226b140ed12b97492116fb03a8790349a","source":{"kind":"arxiv","id":"2111.09734","version":1},"attestation_state":"computed","paper":{"title":"ClipCap: CLIP Prefix for Image Captioning","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Amir Hertz, Amit H. Bermano, Ron Mokady","submitted_at":"2021-11-18T14:49:15Z","abstract_excerpt":"Image captioning is a fundamental task in vision-language understanding, where the model predicts a textual informative caption to a given input image. In this paper, we present a simple approach to address this task. We use CLIP encoding as a prefix to the caption, by employing a simple mapping network, and then fine-tunes a language model to generate the image captions. The recently proposed CLIP model contains rich semantic features which were trained with textual context, making it best for vision-language perception. Our key idea is that together with a pre-trained language model (GPT2), "},"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":"2111.09734","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CV","submitted_at":"2021-11-18T14:49:15Z","cross_cats_sorted":[],"title_canon_sha256":"c2c4134138739a9737c29b2b0b1fe311c1124f104c842f11210de2dccc9846d0","abstract_canon_sha256":"f013ae48f7969453e4c9f3d43dd99453ed01fd855c90ede5b84ac6c28e1745c2"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T03:33:14.847739Z","signature_b64":"Q5jFbqnQPFFBo6KvwrG/qYWjWqvz9t0ZQkN2Twey7JdTtSHBnfd3jzfJgsIXYUyrT3DbLtneByv1l5oKqR9WDw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"b1363424fd029d547c0374a1c904094226b140ed12b97492116fb03a8790349a","last_reissued_at":"2026-07-05T03:33:14.847169Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T03:33:14.847169Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"ClipCap: CLIP Prefix for Image Captioning","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Amir Hertz, Amit H. Bermano, Ron Mokady","submitted_at":"2021-11-18T14:49:15Z","abstract_excerpt":"Image captioning is a fundamental task in vision-language understanding, where the model predicts a textual informative caption to a given input image. In this paper, we present a simple approach to address this task. We use CLIP encoding as a prefix to the caption, by employing a simple mapping network, and then fine-tunes a language model to generate the image captions. The recently proposed CLIP model contains rich semantic features which were trained with textual context, making it best for vision-language perception. Our key idea is that together with a pre-trained language model (GPT2), "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2111.09734","kind":"arxiv","version":1},"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/2111.09734/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":"2111.09734","created_at":"2026-07-05T03:33:14.847252+00:00"},{"alias_kind":"arxiv_version","alias_value":"2111.09734v1","created_at":"2026-07-05T03:33:14.847252+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2111.09734","created_at":"2026-07-05T03:33:14.847252+00:00"},{"alias_kind":"pith_short_12","alias_value":"WE3DIJH5AKOV","created_at":"2026-07-05T03:33:14.847252+00:00"},{"alias_kind":"pith_short_16","alias_value":"WE3DIJH5AKOVI7AD","created_at":"2026-07-05T03:33:14.847252+00:00"},{"alias_kind":"pith_short_8","alias_value":"WE3DIJH5","created_at":"2026-07-05T03:33:14.847252+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":21,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.22766","citing_title":"READ More than What You See: Reinforcement Learning for Accurate and Coherent Audio Description Generations","ref_index":48,"is_internal_anchor":false},{"citing_arxiv_id":"2606.17213","citing_title":"Revisiting LLM Adaptation for 3D CT Report Generation: A Study of Scaling and Diagnostic Priors","ref_index":25,"is_internal_anchor":false},{"citing_arxiv_id":"2606.13288","citing_title":"Cross-Modal Masked Compositional Concept Modeling for Enhancing Visio-Linguistic Compositionality","ref_index":30,"is_internal_anchor":false},{"citing_arxiv_id":"2606.11853","citing_title":"Task-Aware Structured Memory for Dynamic Multi-modal In-Context Learning","ref_index":122,"is_internal_anchor":false},{"citing_arxiv_id":"2606.11074","citing_title":"Modeling Complex Behaviors: Multi-Personality Composition and Dynamic Switching in Vision-Language Models","ref_index":86,"is_internal_anchor":false},{"citing_arxiv_id":"2605.31246","citing_title":"BadBone: Backdoor Attacks Against Backbone Models in Visual Prompt Learning","ref_index":50,"is_internal_anchor":false},{"citing_arxiv_id":"2605.29836","citing_title":"CB-SLICE: Concept-Based Interpretable Error Slice Discovery","ref_index":6,"is_internal_anchor":false},{"citing_arxiv_id":"2601.09298","citing_title":"Multi-Modal LLM based Image Captioning in ICT: Bridging the Gap Between General and Industry Domain","ref_index":24,"is_internal_anchor":false},{"citing_arxiv_id":"2204.00598","citing_title":"Socratic Models: Composing Zero-Shot Multimodal Reasoning with Language","ref_index":45,"is_internal_anchor":false},{"citing_arxiv_id":"2309.08532","citing_title":"EvoPrompt: Connecting LLMs with Evolutionary Algorithms Yields Powerful Prompt Optimizers","ref_index":112,"is_internal_anchor":false},{"citing_arxiv_id":"2304.15010","citing_title":"LLaMA-Adapter V2: Parameter-Efficient Visual Instruction Model","ref_index":46,"is_internal_anchor":false},{"citing_arxiv_id":"2303.16199","citing_title":"LLaMA-Adapter: Efficient Fine-tuning of Language Models with Zero-init Attention","ref_index":146,"is_internal_anchor":false},{"citing_arxiv_id":"2605.13080","citing_title":"Learning to See What You Need: Gaze Attention for Multimodal Large Language Models","ref_index":16,"is_internal_anchor":false},{"citing_arxiv_id":"2210.08402","citing_title":"LAION-5B: An open large-scale dataset for training next generation image-text models","ref_index":52,"is_internal_anchor":false},{"citing_arxiv_id":"2605.09902","citing_title":"Adversarial Attacks Against MLLMs via Progressive Resolution Processing and Adaptive Feature Alignment","ref_index":6,"is_internal_anchor":false},{"citing_arxiv_id":"2204.14198","citing_title":"Flamingo: a Visual Language Model for Few-Shot Learning","ref_index":79,"is_internal_anchor":false},{"citing_arxiv_id":"2604.10132","citing_title":"Semantic Manipulation Localization","ref_index":32,"is_internal_anchor":false},{"citing_arxiv_id":"2604.09442","citing_title":"UIPress: Bringing Optical Token Compression to UI-to-Code Generation","ref_index":31,"is_internal_anchor":false},{"citing_arxiv_id":"2604.04608","citing_title":"Beyond Semantics: Uncovering the Physics of Fakes via Universal Physical Descriptors for Cross-Modal Synthetic Detection","ref_index":18,"is_internal_anchor":false},{"citing_arxiv_id":"2604.05583","citing_title":"WRF4CIR: Weight-Regularized Fine-Tuning Network for Composed Image Retrieval","ref_index":46,"is_internal_anchor":false},{"citing_arxiv_id":"2604.04473","citing_title":"Beyond Standard Benchmarks: A Systematic Audit of Vision-Language Model's Robustness to Natural Semantic Variation Across Diverse Tasks","ref_index":21,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/WE3DIJH5AKOVI7ADOSQ4SBAJII","json":"https://pith.science/pith/WE3DIJH5AKOVI7ADOSQ4SBAJII.json","graph_json":"https://pith.science/api/pith-number/WE3DIJH5AKOVI7ADOSQ4SBAJII/graph.json","events_json":"https://pith.science/api/pith-number/WE3DIJH5AKOVI7ADOSQ4SBAJII/events.json","paper":"https://pith.science/paper/WE3DIJH5"},"agent_actions":{"view_html":"https://pith.science/pith/WE3DIJH5AKOVI7ADOSQ4SBAJII","download_json":"https://pith.science/pith/WE3DIJH5AKOVI7ADOSQ4SBAJII.json","view_paper":"https://pith.science/paper/WE3DIJH5","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2111.09734&json=true","fetch_graph":"https://pith.science/api/pith-number/WE3DIJH5AKOVI7ADOSQ4SBAJII/graph.json","fetch_events":"https://pith.science/api/pith-number/WE3DIJH5AKOVI7ADOSQ4SBAJII/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/WE3DIJH5AKOVI7ADOSQ4SBAJII/action/timestamp_anchor","attest_storage":"https://pith.science/pith/WE3DIJH5AKOVI7ADOSQ4SBAJII/action/storage_attestation","attest_author":"https://pith.science/pith/WE3DIJH5AKOVI7ADOSQ4SBAJII/action/author_attestation","sign_citation":"https://pith.science/pith/WE3DIJH5AKOVI7ADOSQ4SBAJII/action/citation_signature","submit_replication":"https://pith.science/pith/WE3DIJH5AKOVI7ADOSQ4SBAJII/action/replication_record"}},"created_at":"2026-07-05T03:33:14.847252+00:00","updated_at":"2026-07-05T03:33:14.847252+00:00"}