{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2026:TY22YXZ6UOHBRZZESU5YXAKDVR","short_pith_number":"pith:TY22YXZ6","canonical_record":{"source":{"id":"2608.12781","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CV","submitted_at":"2026-08-13T03:44:45Z","cross_cats_sorted":[],"title_canon_sha256":"a060926bf3bbc18b01141cab5f79b6e0ba9f6644dbacaf8cd13fcf82bda31b5b","abstract_canon_sha256":"fad2ffbbb24dfce618ae82ddf7d20869c74de1580dbb8794778256959a2dc9e8"},"schema_version":"1.0"},"canonical_sha256":"9e35ac5f3ea38e18e724953b8b8143ac70e5b8d50881af440d0654976cf7da11","source":{"kind":"arxiv","id":"2608.12781","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2608.12781","created_at":"2026-08-14T00:47:56Z"},{"alias_kind":"arxiv_version","alias_value":"2608.12781v1","created_at":"2026-08-14T00:47:56Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2608.12781","created_at":"2026-08-14T00:47:56Z"},{"alias_kind":"pith_short_12","alias_value":"TY22YXZ6UOHB","created_at":"2026-08-14T00:47:56Z"},{"alias_kind":"pith_short_16","alias_value":"TY22YXZ6UOHBRZZE","created_at":"2026-08-14T00:47:56Z"},{"alias_kind":"pith_short_8","alias_value":"TY22YXZ6","created_at":"2026-08-14T00:47:56Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2026:TY22YXZ6UOHBRZZESU5YXAKDVR","target":"record","payload":{"canonical_record":{"source":{"id":"2608.12781","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CV","submitted_at":"2026-08-13T03:44:45Z","cross_cats_sorted":[],"title_canon_sha256":"a060926bf3bbc18b01141cab5f79b6e0ba9f6644dbacaf8cd13fcf82bda31b5b","abstract_canon_sha256":"fad2ffbbb24dfce618ae82ddf7d20869c74de1580dbb8794778256959a2dc9e8"},"schema_version":"1.0"},"canonical_sha256":"9e35ac5f3ea38e18e724953b8b8143ac70e5b8d50881af440d0654976cf7da11","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-08-14T00:47:56.678234Z","signature_b64":"wt+ESKkKH2F+u8YPkgI/LBKIxxF4kc1m/Ea5b8cY+aivEoTH9FEUnbuOezyrIBczXp+MwDChSTCrxtB+nYAYAA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"9e35ac5f3ea38e18e724953b8b8143ac70e5b8d50881af440d0654976cf7da11","last_reissued_at":"2026-08-14T00:47:56.656026Z","signature_status":"signed_v1","first_computed_at":"2026-08-14T00:47:56.656026Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2608.12781","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":"2026-08-14T00:47:56Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"rL+l651KybXcGWhXqTU5pSmoUuye7yywb88VbyIBnQ0sJALtgzsinnhqzgHqdn6WSX+M+upYfDZCdxPHPUvUDg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-19T20:51:16.940555Z"},"content_sha256":"9e55de9a25e1cc3baf35613f3ec92aa58f7f6956ffaf06792da967d7422f26ed","schema_version":"1.0","event_id":"sha256:9e55de9a25e1cc3baf35613f3ec92aa58f7f6956ffaf06792da967d7422f26ed"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2026:TY22YXZ6UOHBRZZESU5YXAKDVR","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Beyond Correctness: Benchmarking and Aligning Response Behaviors in Hybrid-Thinking MLLMs","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Fengyuan Lu, Guoqing Ma, Han Hu, Hongming Yang, Marito Li, Nan Qiao, Qiming Peng, Saiyong Yang, Shangpin Peng, Songyang Zhang, Weinong Wang, Xinming Wang, Xu-Yao Zhang, Yansong Lin, Yonglong Tian, Zheng Ruan","submitted_at":"2026-08-13T03:44:45Z","abstract_excerpt":"Hybrid-thinking multimodal large language models (MLLMs) allow a single model to alternate between deliberative thinking and latency-efficient non-thinking inference. Although these modes differ in reasoning budget, their delivered responses should satisfy the same user-facing standard. Correctness alone may not characterize this response quality; we therefore evaluate task accuracy and response-pattern failures as complementary outcomes. We study this gap through \\textbf{response-pattern alignment}: whether thinking and non-thinking interfaces preserve acceptable final-response behavior. We i"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2608.12781","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/2608.12781/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"},"verdict_id":null},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-08-14T00:47:56Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"bj/+8VOGpizCCid9eXSnZw3afP8hIbcRdQdS0hw9eH/bRbE4Cn7qgDKGU76W7PLiQ7WACbqv/aYvnGZJlipLAA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-19T20:51:16.940938Z"},"content_sha256":"5ad057070e0169e5a459c6d5e6922461165c47b8c7f002fb7f0f15fc17488004","schema_version":"1.0","event_id":"sha256:5ad057070e0169e5a459c6d5e6922461165c47b8c7f002fb7f0f15fc17488004"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/TY22YXZ6UOHBRZZESU5YXAKDVR/bundle.json","state_url":"https://pith.science/pith/TY22YXZ6UOHBRZZESU5YXAKDVR/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/TY22YXZ6UOHBRZZESU5YXAKDVR/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":"2026-08-19T20:51:16Z","links":{"resolver":"https://pith.science/pith/TY22YXZ6UOHBRZZESU5YXAKDVR","bundle":"https://pith.science/pith/TY22YXZ6UOHBRZZESU5YXAKDVR/bundle.json","state":"https://pith.science/pith/TY22YXZ6UOHBRZZESU5YXAKDVR/state.json","well_known_bundle":"https://pith.science/.well-known/pith/TY22YXZ6UOHBRZZESU5YXAKDVR/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:TY22YXZ6UOHBRZZESU5YXAKDVR","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":"fad2ffbbb24dfce618ae82ddf7d20869c74de1580dbb8794778256959a2dc9e8","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CV","submitted_at":"2026-08-13T03:44:45Z","title_canon_sha256":"a060926bf3bbc18b01141cab5f79b6e0ba9f6644dbacaf8cd13fcf82bda31b5b"},"schema_version":"1.0","source":{"id":"2608.12781","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2608.12781","created_at":"2026-08-14T00:47:56Z"},{"alias_kind":"arxiv_version","alias_value":"2608.12781v1","created_at":"2026-08-14T00:47:56Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2608.12781","created_at":"2026-08-14T00:47:56Z"},{"alias_kind":"pith_short_12","alias_value":"TY22YXZ6UOHB","created_at":"2026-08-14T00:47:56Z"},{"alias_kind":"pith_short_16","alias_value":"TY22YXZ6UOHBRZZE","created_at":"2026-08-14T00:47:56Z"},{"alias_kind":"pith_short_8","alias_value":"TY22YXZ6","created_at":"2026-08-14T00:47:56Z"}],"graph_snapshots":[{"event_id":"sha256:5ad057070e0169e5a459c6d5e6922461165c47b8c7f002fb7f0f15fc17488004","target":"graph","created_at":"2026-08-14T00:47:56Z","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":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"integrity":{"available":true,"clean":true,"detectors_run":[],"endpoint":"/pith/2608.12781/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Hybrid-thinking multimodal large language models (MLLMs) allow a single model to alternate between deliberative thinking and latency-efficient non-thinking inference. Although these modes differ in reasoning budget, their delivered responses should satisfy the same user-facing standard. Correctness alone may not characterize this response quality; we therefore evaluate task accuracy and response-pattern failures as complementary outcomes. We study this gap through \\textbf{response-pattern alignment}: whether thinking and non-thinking interfaces preserve acceptable final-response behavior. We i","authors_text":"Fengyuan Lu, Guoqing Ma, Han Hu, Hongming Yang, Marito Li, Nan Qiao, Qiming Peng, Saiyong Yang, Shangpin Peng, Songyang Zhang, Weinong Wang, Xinming Wang, Xu-Yao Zhang, Yansong Lin, Yonglong Tian, Zheng Ruan","cross_cats":[],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CV","submitted_at":"2026-08-13T03:44:45Z","title":"Beyond Correctness: Benchmarking and Aligning Response Behaviors in Hybrid-Thinking MLLMs"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2608.12781","kind":"arxiv","version":1},"verdict":{"created_at":null,"id":null,"model_set":{},"one_line_summary":"","pipeline_version":null,"pith_extraction_headline":"","strongest_claim":"","weakest_assumption":""}},"verdict_id":null}}],"author_attestations":[],"timestamp_anchors":[],"storage_attestations":[],"citation_signatures":[],"replication_records":[],"corrections":[],"mirror_hints":[],"record_created":{"event_id":"sha256:9e55de9a25e1cc3baf35613f3ec92aa58f7f6956ffaf06792da967d7422f26ed","target":"record","created_at":"2026-08-14T00:47:56Z","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":"fad2ffbbb24dfce618ae82ddf7d20869c74de1580dbb8794778256959a2dc9e8","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CV","submitted_at":"2026-08-13T03:44:45Z","title_canon_sha256":"a060926bf3bbc18b01141cab5f79b6e0ba9f6644dbacaf8cd13fcf82bda31b5b"},"schema_version":"1.0","source":{"id":"2608.12781","kind":"arxiv","version":1}},"canonical_sha256":"9e35ac5f3ea38e18e724953b8b8143ac70e5b8d50881af440d0654976cf7da11","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"9e35ac5f3ea38e18e724953b8b8143ac70e5b8d50881af440d0654976cf7da11","first_computed_at":"2026-08-14T00:47:56.656026Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-08-14T00:47:56.656026Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"wt+ESKkKH2F+u8YPkgI/LBKIxxF4kc1m/Ea5b8cY+aivEoTH9FEUnbuOezyrIBczXp+MwDChSTCrxtB+nYAYAA==","signature_status":"signed_v1","signed_at":"2026-08-14T00:47:56.678234Z","signed_message":"canonical_sha256_bytes"},"source_id":"2608.12781","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:9e55de9a25e1cc3baf35613f3ec92aa58f7f6956ffaf06792da967d7422f26ed","sha256:5ad057070e0169e5a459c6d5e6922461165c47b8c7f002fb7f0f15fc17488004"],"state_sha256":"9fed07bbcff1c6e40d7a039c7311ffb14e65f403b0136f36e0dec364078ff2e1"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"9S77Y0KdeNcAnlPnEljZk6MW2NWeQjyK87XOhR9IDmSJ16JiVbsW4B9UiC8Ouah+QvDknD3HjhlHcnnwEcuMDQ==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-19T20:51:16.943412Z","bundle_sha256":"5d670acd66cd44324bfcd7a80b3ef592da5223935b82d0dc5b38f20a52723387"}}