{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:MGC4C3F3QELRLJEINJWQI5K6EP","short_pith_number":"pith:MGC4C3F3","schema_version":"1.0","canonical_sha256":"6185c16cbb811715a4886a6d04755e23f378544c664502196b04d6c996c2db96","source":{"kind":"arxiv","id":"2507.15846","version":3},"attestation_state":"computed","paper":{"title":"GUI-G$^2$: Gaussian Reward Modeling for GUI Grounding","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.CL","cs.CV","cs.HC"],"primary_cat":"cs.LG","authors_text":"Changhua Meng, Fei Tang, Jun Xiao, Shuheng Shen, Weiming Lu, Wenqi Zhang, Wen Wang, Xuyang Liu, Yongliang Shen, Yueting Zhuang, Zhangxuan Gu, Zhengxi Lu","submitted_at":"2025-07-21T17:53:42Z","abstract_excerpt":"Graphical User Interface (GUI) grounding maps natural language instructions to precise interface locations for autonomous interaction. Current reinforcement learning approaches use binary rewards that treat elements as hit-or-miss targets, creating sparse signals that ignore the continuous nature of spatial interactions. Motivated by human clicking behavior that naturally forms Gaussian distributions centered on target elements, we introduce GUI Gaussian Grounding Rewards (GUI-G$^2$), a principled reward framework that models GUI elements as continuous Gaussian distributions across the interfa"},"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":"2507.15846","kind":"arxiv","version":3},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2025-07-21T17:53:42Z","cross_cats_sorted":["cs.AI","cs.CL","cs.CV","cs.HC"],"title_canon_sha256":"fed5e98fc8e55f68d65d47d556a97c8b1d893e1b5311516956d06b1430d8d841","abstract_canon_sha256":"2df047e480e6c79146d2549a15ab0d15b1941c3f5f9571075d4a6ba54a60973b"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:44:09.625295Z","signature_b64":"UfdR4mSjgKEhGES/ihXsE4ihfjxSbUyX/OMXzYWs3WMKrfI2UMxvk8AJNwm7+ZYAj6GygOarIOURLTrSjCkOBw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"6185c16cbb811715a4886a6d04755e23f378544c664502196b04d6c996c2db96","last_reissued_at":"2026-07-05T11:44:09.624745Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:44:09.624745Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"GUI-G$^2$: Gaussian Reward Modeling for GUI Grounding","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.CL","cs.CV","cs.HC"],"primary_cat":"cs.LG","authors_text":"Changhua Meng, Fei Tang, Jun Xiao, Shuheng Shen, Weiming Lu, Wenqi Zhang, Wen Wang, Xuyang Liu, Yongliang Shen, Yueting Zhuang, Zhangxuan Gu, Zhengxi Lu","submitted_at":"2025-07-21T17:53:42Z","abstract_excerpt":"Graphical User Interface (GUI) grounding maps natural language instructions to precise interface locations for autonomous interaction. Current reinforcement learning approaches use binary rewards that treat elements as hit-or-miss targets, creating sparse signals that ignore the continuous nature of spatial interactions. Motivated by human clicking behavior that naturally forms Gaussian distributions centered on target elements, we introduce GUI Gaussian Grounding Rewards (GUI-G$^2$), a principled reward framework that models GUI elements as continuous Gaussian distributions across the interfa"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2507.15846","kind":"arxiv","version":3},"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/2507.15846/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":"2507.15846","created_at":"2026-07-05T11:44:09.624826+00:00"},{"alias_kind":"arxiv_version","alias_value":"2507.15846v3","created_at":"2026-07-05T11:44:09.624826+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2507.15846","created_at":"2026-07-05T11:44:09.624826+00:00"},{"alias_kind":"pith_short_12","alias_value":"MGC4C3F3QELR","created_at":"2026-07-05T11:44:09.624826+00:00"},{"alias_kind":"pith_short_16","alias_value":"MGC4C3F3QELRLJEI","created_at":"2026-07-05T11:44:09.624826+00:00"},{"alias_kind":"pith_short_8","alias_value":"MGC4C3F3","created_at":"2026-07-05T11:44:09.624826+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":14,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.10522","citing_title":"GUI-AC: Enhancing Continual Learning in GUI Agents","ref_index":23,"is_internal_anchor":false},{"citing_arxiv_id":"2605.28629","citing_title":"Mobile-Aptus: Confidence-Driven Proactive and Robust Interaction in MLLM-based Mobile-Using Agents","ref_index":19,"is_internal_anchor":false},{"citing_arxiv_id":"2605.30884","citing_title":"GUI-C$^2$: Coarse-to-Fine GUI Grounding via Difficulty-Aware Reinforcement Learning","ref_index":32,"is_internal_anchor":false},{"citing_arxiv_id":"2509.07553","citing_title":"VeriOS: Query-Driven Proactive Human-Agent-GUI Interaction for Trustworthy OS Agents","ref_index":45,"is_internal_anchor":false},{"citing_arxiv_id":"2509.21982","citing_title":"RISK: A Framework for GUI Agents in E-commerce Risk Management","ref_index":20,"is_internal_anchor":false},{"citing_arxiv_id":"2603.26041","citing_title":"Where and How to Prune: An Empirical Study of Visual Token Pruning for GUI Agent Navigation","ref_index":23,"is_internal_anchor":false},{"citing_arxiv_id":"2605.12501","citing_title":"Covering Human Action Space for Computer Use: Data Synthesis and Benchmark","ref_index":20,"is_internal_anchor":false},{"citing_arxiv_id":"2604.27955","citing_title":"GUI Agents with Reinforcement Learning: Toward Digital Inhabitants","ref_index":67,"is_internal_anchor":false},{"citing_arxiv_id":"2604.24348","citing_title":"OS-SPEAR: A Toolkit for the Safety, Performance,Efficiency, and Robustness Analysis of OS Agents","ref_index":25,"is_internal_anchor":false},{"citing_arxiv_id":"2605.06664","citing_title":"BAMI: Training-Free Bias Mitigation in GUI Grounding","ref_index":26,"is_internal_anchor":false},{"citing_arxiv_id":"2604.21268","citing_title":"Measure Twice, Click Once: Co-evolving Proposer and Visual Critic via Reinforcement Learning for GUI Grounding","ref_index":66,"is_internal_anchor":false},{"citing_arxiv_id":"2604.13019","citing_title":"PrecisionCUA: Iterative Visual Refinement for Pixel-Precise Cursor Grounding in Code Editors","ref_index":12,"is_internal_anchor":false},{"citing_arxiv_id":"2604.09442","citing_title":"UIPress: Bringing Optical Token Compression to UI-to-Code Generation","ref_index":46,"is_internal_anchor":false},{"citing_arxiv_id":"2604.13531","citing_title":"RiskWebWorld: A Realistic Interactive Benchmark for GUI Agents in E-commerce Risk Management","ref_index":43,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/MGC4C3F3QELRLJEINJWQI5K6EP","json":"https://pith.science/pith/MGC4C3F3QELRLJEINJWQI5K6EP.json","graph_json":"https://pith.science/api/pith-number/MGC4C3F3QELRLJEINJWQI5K6EP/graph.json","events_json":"https://pith.science/api/pith-number/MGC4C3F3QELRLJEINJWQI5K6EP/events.json","paper":"https://pith.science/paper/MGC4C3F3"},"agent_actions":{"view_html":"https://pith.science/pith/MGC4C3F3QELRLJEINJWQI5K6EP","download_json":"https://pith.science/pith/MGC4C3F3QELRLJEINJWQI5K6EP.json","view_paper":"https://pith.science/paper/MGC4C3F3","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2507.15846&json=true","fetch_graph":"https://pith.science/api/pith-number/MGC4C3F3QELRLJEINJWQI5K6EP/graph.json","fetch_events":"https://pith.science/api/pith-number/MGC4C3F3QELRLJEINJWQI5K6EP/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/MGC4C3F3QELRLJEINJWQI5K6EP/action/timestamp_anchor","attest_storage":"https://pith.science/pith/MGC4C3F3QELRLJEINJWQI5K6EP/action/storage_attestation","attest_author":"https://pith.science/pith/MGC4C3F3QELRLJEINJWQI5K6EP/action/author_attestation","sign_citation":"https://pith.science/pith/MGC4C3F3QELRLJEINJWQI5K6EP/action/citation_signature","submit_replication":"https://pith.science/pith/MGC4C3F3QELRLJEINJWQI5K6EP/action/replication_record"}},"created_at":"2026-07-05T11:44:09.624826+00:00","updated_at":"2026-07-05T11:44:09.624826+00:00"}