{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2026:BBPZHPOC6RTCXFWFXO5RH5O4RX","short_pith_number":"pith:BBPZHPOC","canonical_record":{"source":{"id":"2605.14311","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-05-14T03:23:44Z","cross_cats_sorted":["cs.AI","cs.HC"],"title_canon_sha256":"0106bb1267b78a413da48e8a02f0dfa02c326bc3222d64832b3aa12f85189bd6","abstract_canon_sha256":"21c60aaafb05789cc955b15ccecff7d5f39f3b40988f6ed42c9cdf6e3cb0af26"},"schema_version":"1.0"},"canonical_sha256":"085f93bdc2f4662b96c5bbbb13f5dc8dfd850cdeace4e215984f2cc34b831038","source":{"kind":"arxiv","id":"2605.14311","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2605.14311","created_at":"2026-05-17T23:39:09Z"},{"alias_kind":"arxiv_version","alias_value":"2605.14311v1","created_at":"2026-05-17T23:39:09Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.14311","created_at":"2026-05-17T23:39:09Z"},{"alias_kind":"pith_short_12","alias_value":"BBPZHPOC6RTC","created_at":"2026-05-18T12:33:37Z"},{"alias_kind":"pith_short_16","alias_value":"BBPZHPOC6RTCXFWF","created_at":"2026-05-18T12:33:37Z"},{"alias_kind":"pith_short_8","alias_value":"BBPZHPOC","created_at":"2026-05-18T12:33:37Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2026:BBPZHPOC6RTCXFWFXO5RH5O4RX","target":"record","payload":{"canonical_record":{"source":{"id":"2605.14311","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-05-14T03:23:44Z","cross_cats_sorted":["cs.AI","cs.HC"],"title_canon_sha256":"0106bb1267b78a413da48e8a02f0dfa02c326bc3222d64832b3aa12f85189bd6","abstract_canon_sha256":"21c60aaafb05789cc955b15ccecff7d5f39f3b40988f6ed42c9cdf6e3cb0af26"},"schema_version":"1.0"},"canonical_sha256":"085f93bdc2f4662b96c5bbbb13f5dc8dfd850cdeace4e215984f2cc34b831038","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-17T23:39:09.994422Z","signature_b64":"v1D7fcJanfDFTumQCPk/ZJuncMr+s+RomLw8A01gPSvpE3BNyWs9yidbVrq7/7hAZw8H0OW+ICsTAsr4kLhWDw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"085f93bdc2f4662b96c5bbbb13f5dc8dfd850cdeace4e215984f2cc34b831038","last_reissued_at":"2026-05-17T23:39:09.993776Z","signature_status":"signed_v1","first_computed_at":"2026-05-17T23:39:09.993776Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2605.14311","source_version":1,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-05-17T23:39:09Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"c6WBAxo2IrotniadbSBmidKm6CI/KebPGES6/6G5/QMuDnNSQ/Z6iExOsrqiQ9a7qR4iMfkMD/qxPWD5hioIAw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-05-30T14:23:04.152514Z"},"content_sha256":"6547e358fdbd03b22a1da4dae25b657a7e84d9f6fcec685cf7b8c0f37a01a110","schema_version":"1.0","event_id":"sha256:6547e358fdbd03b22a1da4dae25b657a7e84d9f6fcec685cf7b8c0f37a01a110"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2026:BBPZHPOC6RTCXFWFXO5RH5O4RX","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Beyond Binary: Reframing GUI Critique as Continuous Semantic Alignment","license":"http://creativecommons.org/licenses/by/4.0/","headline":"GUI critique improves when reframed as continuous semantic alignment in a shared affordance space rather than binary classification.","cross_cats":["cs.AI","cs.HC"],"primary_cat":"cs.LG","authors_text":"Anan Du, Chongyang Zhang, Jian Luan, Pei Fu, Ruoceng Zhang, Shaojie Zhang, Xiuwen Xi, Yuchen Sun, Zhenbo Luo","submitted_at":"2026-05-14T03:23:44Z","abstract_excerpt":"Test-Time Scaling (TTS), which samples multiple candidate actions and ranks them via a Critic Model, has emerged as a promising paradigm for generalist GUI agents. Its efficacy thus hinges on the critic's fine-grained ranking ability. However, existing GUI critic models uniformly adopt binary classification. Our motivational analysis of these models exposes a severe entanglement: scores for valid actions and plausible-but-invalid distractors become indistinguishable. We attribute this failure to two structural defects: Affordance Collapse--the hierarchical affordance space is compressed into 0"},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"BBCritic-3B, trained without any extra annotation, outperforms 7B-parameter SOTA binary models and demonstrates strong zero-shot transferability across platforms and tasks.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"The Functional Equivalence Hypothesis that a shared continuous Affordance Space can recover the hierarchical structure flattened by binary labels; this is invoked to justify the two-stage contrastive objective but is not independently validated in the provided abstract.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"BBCritic uses contrastive learning to align GUI actions in a continuous affordance space, outperforming larger binary critic models on a new four-level hierarchical benchmark while enabling zero-shot transfer.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"GUI critique improves when reframed as continuous semantic alignment in a shared affordance space rather than binary classification.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"41245c163d1f9d45003e588d4d724de7c65a092e3f87e17b957cba659d5a28db"},"source":{"id":"2605.14311","kind":"arxiv","version":1},"verdict":{"id":"beb4e7df-8eae-453c-8b45-decf84a6d5e2","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-15T01:54:24.729643Z","strongest_claim":"BBCritic-3B, trained without any extra annotation, outperforms 7B-parameter SOTA binary models and demonstrates strong zero-shot transferability across platforms and tasks.","one_line_summary":"BBCritic uses contrastive learning to align GUI actions in a continuous affordance space, outperforming larger binary critic models on a new four-level hierarchical benchmark while enabling zero-shot transfer.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"The Functional Equivalence Hypothesis that a shared continuous Affordance Space can recover the hierarchical structure flattened by binary labels; this is invoked to justify the two-stage contrastive objective but is not independently validated in the provided abstract.","pith_extraction_headline":"GUI critique improves when reframed as continuous semantic alignment in a shared affordance space rather than binary classification."},"references":{"count":93,"sample":[{"doi":"","year":2025,"title":"Proceedings of the 2025 Conference on Empirical Methods in Natural Language Processing , pages=","work_id":"b961c8d2-0b87-4e53-b6c2-b0c0514e100e","ref_index":4,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":null,"title":"Advances in Neural Information Processing Systems , volume=","work_id":"f95fb7e2-29a6-4f31-88f2-3f1ba5b63f25","ref_index":10,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":null,"title":"arXiv e-prints , pages=","work_id":"231ade17-1155-4046-8cb9-73cf7a223698","ref_index":12,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2024,"title":"Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing , pages=","work_id":"e342ad04-078a-41eb-8767-9e7645fd2bd1","ref_index":13,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":1999,"title":"Affordance, conventions, and design , author=. interactions , volume=. 1999 , publisher=","work_id":"779b0e65-7adb-4209-bb74-797c07e4bac3","ref_index":20,"cited_arxiv_id":"","is_internal_anchor":false}],"resolved_work":93,"snapshot_sha256":"3c33762321439455588e7e4fa4586655c19e6a76230dabed99e370c80ea28d3b","internal_anchors":21},"formal_canon":{"evidence_count":2,"snapshot_sha256":"acb660c6e80c287dac3169d91d84c8998d0088fbe4b9ac70ccf3167585968f89"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"},"verdict_id":"beb4e7df-8eae-453c-8b45-decf84a6d5e2"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-05-17T23:39:09Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"2Oyo6PxeqcdDyY2lElbu6XBzroM5hr3cX2Qop0UCn01KZOPC+2G6Gj+GKu/jlDNmAlQ/C3s+ZljRW3GxeJvFDA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-05-30T14:23:04.153351Z"},"content_sha256":"e5bd4ecb230321fbd24d61f16f014bf02a631d7278d87a9e4e54abc93676ad0c","schema_version":"1.0","event_id":"sha256:e5bd4ecb230321fbd24d61f16f014bf02a631d7278d87a9e4e54abc93676ad0c"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/BBPZHPOC6RTCXFWFXO5RH5O4RX/bundle.json","state_url":"https://pith.science/pith/BBPZHPOC6RTCXFWFXO5RH5O4RX/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/BBPZHPOC6RTCXFWFXO5RH5O4RX/bundle.json","status":"primary"}],"public_keys":[{"key_id":"pith-v1-2026-05","algorithm":"ed25519","format":"raw","public_key_b64":"stVStoiQhXFxp4s2pdzPNoqVNBMojDU/fJ2db5S3CbM=","public_key_hex":"b2d552b68890857171a78b36a5dccf368a953413288c353f7c9d9d6f94b709b3","fingerprint_sha256_b32_first128bits":"RVFV5Z2OI2J3ZUO7ERDEBCYNKS","fingerprint_sha256_hex":"8d4b5ee74e4693bcd1df2446408b0d54","rotates_at":null,"url":"https://pith.science/pith-signing-key.json","notes":"Pith uses this Ed25519 key to sign canonical record SHA-256 digests. Verify with: ed25519_verify(public_key, message=canonical_sha256_bytes, signature=base64decode(signature_b64))."}],"merge_version":"pith-open-graph-merge-v1","built_at":"2026-05-30T14:23:04Z","links":{"resolver":"https://pith.science/pith/BBPZHPOC6RTCXFWFXO5RH5O4RX","bundle":"https://pith.science/pith/BBPZHPOC6RTCXFWFXO5RH5O4RX/bundle.json","state":"https://pith.science/pith/BBPZHPOC6RTCXFWFXO5RH5O4RX/state.json","well_known_bundle":"https://pith.science/.well-known/pith/BBPZHPOC6RTCXFWFXO5RH5O4RX/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:BBPZHPOC6RTCXFWFXO5RH5O4RX","merge_version":"pith-open-graph-merge-v1","event_count":2,"valid_event_count":2,"invalid_event_count":0,"equivocation_count":0,"current":{"canonical_record":{"metadata":{"abstract_canon_sha256":"21c60aaafb05789cc955b15ccecff7d5f39f3b40988f6ed42c9cdf6e3cb0af26","cross_cats_sorted":["cs.AI","cs.HC"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-05-14T03:23:44Z","title_canon_sha256":"0106bb1267b78a413da48e8a02f0dfa02c326bc3222d64832b3aa12f85189bd6"},"schema_version":"1.0","source":{"id":"2605.14311","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2605.14311","created_at":"2026-05-17T23:39:09Z"},{"alias_kind":"arxiv_version","alias_value":"2605.14311v1","created_at":"2026-05-17T23:39:09Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.14311","created_at":"2026-05-17T23:39:09Z"},{"alias_kind":"pith_short_12","alias_value":"BBPZHPOC6RTC","created_at":"2026-05-18T12:33:37Z"},{"alias_kind":"pith_short_16","alias_value":"BBPZHPOC6RTCXFWF","created_at":"2026-05-18T12:33:37Z"},{"alias_kind":"pith_short_8","alias_value":"BBPZHPOC","created_at":"2026-05-18T12:33:37Z"}],"graph_snapshots":[{"event_id":"sha256:e5bd4ecb230321fbd24d61f16f014bf02a631d7278d87a9e4e54abc93676ad0c","target":"graph","created_at":"2026-05-17T23:39:09Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"graph_snapshot":{"author_claims":{"count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","strong_count":0},"builder_version":"pith-number-builder-2026-05-17-v1","claims":{"count":4,"items":[{"attestation":"unclaimed","claim_id":"C1","kind":"strongest_claim","source":"verdict.strongest_claim","status":"machine_extracted","text":"BBCritic-3B, trained without any extra annotation, outperforms 7B-parameter SOTA binary models and demonstrates strong zero-shot transferability across platforms and tasks."},{"attestation":"unclaimed","claim_id":"C2","kind":"weakest_assumption","source":"verdict.weakest_assumption","status":"machine_extracted","text":"The Functional Equivalence Hypothesis that a shared continuous Affordance Space can recover the hierarchical structure flattened by binary labels; this is invoked to justify the two-stage contrastive objective but is not independently validated in the provided abstract."},{"attestation":"unclaimed","claim_id":"C3","kind":"one_line_summary","source":"verdict.one_line_summary","status":"machine_extracted","text":"BBCritic uses contrastive learning to align GUI actions in a continuous affordance space, outperforming larger binary critic models on a new four-level hierarchical benchmark while enabling zero-shot transfer."},{"attestation":"unclaimed","claim_id":"C4","kind":"headline","source":"verdict.pith_extraction.headline","status":"machine_extracted","text":"GUI critique improves when reframed as continuous semantic alignment in a shared affordance space rather than binary classification."}],"snapshot_sha256":"41245c163d1f9d45003e588d4d724de7c65a092e3f87e17b957cba659d5a28db"},"formal_canon":{"evidence_count":2,"snapshot_sha256":"acb660c6e80c287dac3169d91d84c8998d0088fbe4b9ac70ccf3167585968f89"},"paper":{"abstract_excerpt":"Test-Time Scaling (TTS), which samples multiple candidate actions and ranks them via a Critic Model, has emerged as a promising paradigm for generalist GUI agents. Its efficacy thus hinges on the critic's fine-grained ranking ability. However, existing GUI critic models uniformly adopt binary classification. Our motivational analysis of these models exposes a severe entanglement: scores for valid actions and plausible-but-invalid distractors become indistinguishable. We attribute this failure to two structural defects: Affordance Collapse--the hierarchical affordance space is compressed into 0","authors_text":"Anan Du, Chongyang Zhang, Jian Luan, Pei Fu, Ruoceng Zhang, Shaojie Zhang, Xiuwen Xi, Yuchen Sun, Zhenbo Luo","cross_cats":["cs.AI","cs.HC"],"headline":"GUI critique improves when reframed as continuous semantic alignment in a shared affordance space rather than binary classification.","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-05-14T03:23:44Z","title":"Beyond Binary: Reframing GUI Critique as Continuous Semantic Alignment"},"references":{"count":93,"internal_anchors":21,"resolved_work":93,"sample":[{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":4,"title":"Proceedings of the 2025 Conference on Empirical Methods in Natural Language Processing , pages=","work_id":"b961c8d2-0b87-4e53-b6c2-b0c0514e100e","year":2025},{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":10,"title":"Advances in Neural Information Processing Systems , volume=","work_id":"f95fb7e2-29a6-4f31-88f2-3f1ba5b63f25","year":null},{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":12,"title":"arXiv e-prints , pages=","work_id":"231ade17-1155-4046-8cb9-73cf7a223698","year":null},{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":13,"title":"Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing , pages=","work_id":"e342ad04-078a-41eb-8767-9e7645fd2bd1","year":2024},{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":20,"title":"Affordance, conventions, and design , author=. interactions , volume=. 1999 , publisher=","work_id":"779b0e65-7adb-4209-bb74-797c07e4bac3","year":1999}],"snapshot_sha256":"3c33762321439455588e7e4fa4586655c19e6a76230dabed99e370c80ea28d3b"},"source":{"id":"2605.14311","kind":"arxiv","version":1},"verdict":{"created_at":"2026-05-15T01:54:24.729643Z","id":"beb4e7df-8eae-453c-8b45-decf84a6d5e2","model_set":{"reader":"grok-4.3"},"one_line_summary":"BBCritic uses contrastive learning to align GUI actions in a continuous affordance space, outperforming larger binary critic models on a new four-level hierarchical benchmark while enabling zero-shot transfer.","pipeline_version":"pith-pipeline@v0.9.0","pith_extraction_headline":"GUI critique improves when reframed as continuous semantic alignment in a shared affordance space rather than binary classification.","strongest_claim":"BBCritic-3B, trained without any extra annotation, outperforms 7B-parameter SOTA binary models and demonstrates strong zero-shot transferability across platforms and tasks.","weakest_assumption":"The Functional Equivalence Hypothesis that a shared continuous Affordance Space can recover the hierarchical structure flattened by binary labels; this is invoked to justify the two-stage contrastive objective but is not independently validated in the provided abstract."}},"verdict_id":"beb4e7df-8eae-453c-8b45-decf84a6d5e2"}}],"author_attestations":[],"timestamp_anchors":[],"storage_attestations":[],"citation_signatures":[],"replication_records":[],"corrections":[],"mirror_hints":[],"record_created":{"event_id":"sha256:6547e358fdbd03b22a1da4dae25b657a7e84d9f6fcec685cf7b8c0f37a01a110","target":"record","created_at":"2026-05-17T23:39:09Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"attestation_state":"computed","canonical_record":{"metadata":{"abstract_canon_sha256":"21c60aaafb05789cc955b15ccecff7d5f39f3b40988f6ed42c9cdf6e3cb0af26","cross_cats_sorted":["cs.AI","cs.HC"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-05-14T03:23:44Z","title_canon_sha256":"0106bb1267b78a413da48e8a02f0dfa02c326bc3222d64832b3aa12f85189bd6"},"schema_version":"1.0","source":{"id":"2605.14311","kind":"arxiv","version":1}},"canonical_sha256":"085f93bdc2f4662b96c5bbbb13f5dc8dfd850cdeace4e215984f2cc34b831038","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"085f93bdc2f4662b96c5bbbb13f5dc8dfd850cdeace4e215984f2cc34b831038","first_computed_at":"2026-05-17T23:39:09.993776Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-05-17T23:39:09.993776Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"v1D7fcJanfDFTumQCPk/ZJuncMr+s+RomLw8A01gPSvpE3BNyWs9yidbVrq7/7hAZw8H0OW+ICsTAsr4kLhWDw==","signature_status":"signed_v1","signed_at":"2026-05-17T23:39:09.994422Z","signed_message":"canonical_sha256_bytes"},"source_id":"2605.14311","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:6547e358fdbd03b22a1da4dae25b657a7e84d9f6fcec685cf7b8c0f37a01a110","sha256:e5bd4ecb230321fbd24d61f16f014bf02a631d7278d87a9e4e54abc93676ad0c"],"state_sha256":"8fd12dde7743834ef405b445fd25fd1ea4e25381ddcc29a52a852e791a9ef975"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"UrjiYXFaOBMoRVTTb9DH3kXBYurKtij4en+vqEMPhNJdsSTjv+PONK58obY9au2d3HYyMB1cbjb66G94HrqzBw==","signed_message":"bundle_sha256_bytes","signed_at":"2026-05-30T14:23:04.156766Z","bundle_sha256":"1df3a76e8ebbbfd4917ae87e8124c7de2505a99f0a9c22f26250f203acf3f960"}}