{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2023:ILUY5LTU57ZL4OD77JKI2C65SW","short_pith_number":"pith:ILUY5LTU","schema_version":"1.0","canonical_sha256":"42e98eae74eff2be387ffa548d0bdd95adc7dc9b04b180adf1847992ce2e1935","source":{"kind":"arxiv","id":"2305.11854","version":4},"attestation_state":"computed","paper":{"title":"Multimodal Web Navigation with Instruction-Finetuned Foundation Models","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","stat.ML"],"primary_cat":"cs.LG","authors_text":"Aleksandra Faust, Hiroki Furuta, Izzeddin Gur, Kuang-Huei Lee, Ofir Nachum, Shixiang Shane Gu, Yutaka Matsuo","submitted_at":"2023-05-19T17:44:34Z","abstract_excerpt":"The progress of autonomous web navigation has been hindered by the dependence on billions of exploratory interactions via online reinforcement learning, and domain-specific model designs that make it difficult to leverage generalization from rich out-of-domain data. In this work, we study data-driven offline training for web agents with vision-language foundation models. We propose an instruction-following multimodal agent, WebGUM, that observes both webpage screenshots and HTML pages and outputs web navigation actions, such as click and type. WebGUM is trained by jointly finetuning an instruc"},"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":"2305.11854","kind":"arxiv","version":4},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2023-05-19T17:44:34Z","cross_cats_sorted":["cs.AI","stat.ML"],"title_canon_sha256":"1c44c120f0843ffb15c1a06f57c527fa0ebbbd101cfe9b95e4e6f019c81f5b53","abstract_canon_sha256":"2fe5e6b9ee2b739e6a8552b172499cf658eb354b4e1a52a66c1d12ce12f70c8f"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T07:48:46.459857Z","signature_b64":"rXLSaEyDWRby8hf0sXllc4VDLYulmXokaud8RUMk0Hltzt5RkYiJK7GRnBXRkI27YCKGtc+mL1eo3yPI2BKkAA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"42e98eae74eff2be387ffa548d0bdd95adc7dc9b04b180adf1847992ce2e1935","last_reissued_at":"2026-07-05T07:48:46.459345Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T07:48:46.459345Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Multimodal Web Navigation with Instruction-Finetuned Foundation Models","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","stat.ML"],"primary_cat":"cs.LG","authors_text":"Aleksandra Faust, Hiroki Furuta, Izzeddin Gur, Kuang-Huei Lee, Ofir Nachum, Shixiang Shane Gu, Yutaka Matsuo","submitted_at":"2023-05-19T17:44:34Z","abstract_excerpt":"The progress of autonomous web navigation has been hindered by the dependence on billions of exploratory interactions via online reinforcement learning, and domain-specific model designs that make it difficult to leverage generalization from rich out-of-domain data. In this work, we study data-driven offline training for web agents with vision-language foundation models. We propose an instruction-following multimodal agent, WebGUM, that observes both webpage screenshots and HTML pages and outputs web navigation actions, such as click and type. WebGUM is trained by jointly finetuning an instruc"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2305.11854","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/2305.11854/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":"2305.11854","created_at":"2026-07-05T07:48:46.459416+00:00"},{"alias_kind":"arxiv_version","alias_value":"2305.11854v4","created_at":"2026-07-05T07:48:46.459416+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2305.11854","created_at":"2026-07-05T07:48:46.459416+00:00"},{"alias_kind":"pith_short_12","alias_value":"ILUY5LTU57ZL","created_at":"2026-07-05T07:48:46.459416+00:00"},{"alias_kind":"pith_short_16","alias_value":"ILUY5LTU57ZL4OD7","created_at":"2026-07-05T07:48:46.459416+00:00"},{"alias_kind":"pith_short_8","alias_value":"ILUY5LTU","created_at":"2026-07-05T07:48:46.459416+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":13,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2607.00007","citing_title":"BaRA: Budget-constrained and Reliable Web Data Collection Agent","ref_index":18,"is_internal_anchor":false},{"citing_arxiv_id":"2406.12373","citing_title":"WebCanvas: Benchmarking Web Agents in Online Environments","ref_index":5,"is_internal_anchor":false},{"citing_arxiv_id":"2411.18279","citing_title":"Large Language Model-Brained GUI Agents: A Survey","ref_index":156,"is_internal_anchor":false},{"citing_arxiv_id":"2510.22102","citing_title":"Mitigating Coordinate Prediction Bias from Positional Encoding Failures","ref_index":5,"is_internal_anchor":false},{"citing_arxiv_id":"2510.23883","citing_title":"Agentic AI Security: Threats, Defenses, Evaluation, and Open Challenges","ref_index":231,"is_internal_anchor":false},{"citing_arxiv_id":"2503.09572","citing_title":"Plan-and-Act: Improving Planning of Agents for Long-Horizon Tasks","ref_index":6,"is_internal_anchor":false},{"citing_arxiv_id":"2312.13771","citing_title":"AppAgent: Multimodal Agents as Smartphone Users","ref_index":6,"is_internal_anchor":false},{"citing_arxiv_id":"2401.10935","citing_title":"SeeClick: Harnessing GUI Grounding for Advanced Visual GUI Agents","ref_index":75,"is_internal_anchor":false},{"citing_arxiv_id":"2401.05459","citing_title":"Personal LLM Agents: Insights and Survey about the Capability, Efficiency and Security","ref_index":49,"is_internal_anchor":false},{"citing_arxiv_id":"2603.05044","citing_title":"WebFactory: Automated Compression of Foundational Language Intelligence into Grounded Web Agents","ref_index":7,"is_internal_anchor":false},{"citing_arxiv_id":"2403.07718","citing_title":"WorkArena: How Capable Are Web Agents at Solving Common Knowledge Work Tasks?","ref_index":4,"is_internal_anchor":false},{"citing_arxiv_id":"2404.07972","citing_title":"OSWorld: Benchmarking Multimodal Agents for Open-Ended Tasks in Real Computer Environments","ref_index":12,"is_internal_anchor":false},{"citing_arxiv_id":"2604.26020","citing_title":"Training Computer Use Agents to Assess the Usability of Graphical User Interfaces","ref_index":16,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/ILUY5LTU57ZL4OD77JKI2C65SW","json":"https://pith.science/pith/ILUY5LTU57ZL4OD77JKI2C65SW.json","graph_json":"https://pith.science/api/pith-number/ILUY5LTU57ZL4OD77JKI2C65SW/graph.json","events_json":"https://pith.science/api/pith-number/ILUY5LTU57ZL4OD77JKI2C65SW/events.json","paper":"https://pith.science/paper/ILUY5LTU"},"agent_actions":{"view_html":"https://pith.science/pith/ILUY5LTU57ZL4OD77JKI2C65SW","download_json":"https://pith.science/pith/ILUY5LTU57ZL4OD77JKI2C65SW.json","view_paper":"https://pith.science/paper/ILUY5LTU","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2305.11854&json=true","fetch_graph":"https://pith.science/api/pith-number/ILUY5LTU57ZL4OD77JKI2C65SW/graph.json","fetch_events":"https://pith.science/api/pith-number/ILUY5LTU57ZL4OD77JKI2C65SW/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/ILUY5LTU57ZL4OD77JKI2C65SW/action/timestamp_anchor","attest_storage":"https://pith.science/pith/ILUY5LTU57ZL4OD77JKI2C65SW/action/storage_attestation","attest_author":"https://pith.science/pith/ILUY5LTU57ZL4OD77JKI2C65SW/action/author_attestation","sign_citation":"https://pith.science/pith/ILUY5LTU57ZL4OD77JKI2C65SW/action/citation_signature","submit_replication":"https://pith.science/pith/ILUY5LTU57ZL4OD77JKI2C65SW/action/replication_record"}},"created_at":"2026-07-05T07:48:46.459416+00:00","updated_at":"2026-07-05T07:48:46.459416+00:00"}