{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:E264CY2MSQF4FZ7Q5HM4BUTW27","short_pith_number":"pith:E264CY2M","schema_version":"1.0","canonical_sha256":"26bdc1634c940bc2e7f0e9d9c0d276d7e521aced7b4b7990269fc89d627a8187","source":{"kind":"arxiv","id":"2604.20328","version":2},"attestation_state":"computed","paper":{"title":"HyLaR: Hybrid Latent Reasoning with Decoupled Policy Optimization","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"Hybrid latent reasoning interleaves discrete text generation with continuous visual states and optimizes the combination through decoupled policy learning.","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Hao Zhang, Jinwen Luo, Shi-Zhe Chen, Tao Cheng, Yixin Qin, Zheng Wei","submitted_at":"2026-04-22T08:22:23Z","abstract_excerpt":"Chain-of-Thought (CoT) reasoning significantly elevates the complex problem-solving capabilities of multimodal large language models (MLLMs). However, adapting CoT to vision typically discretizes signals to fit LLM inputs, causing early semantic collapse and discarding fine-grained details. While external tools can mitigate this, they introduce a rigid bottleneck, confining reasoning to predefined operations. Although recent latent reasoning paradigms internalize visual states to overcome these limitations, optimizing the resulting hybrid discrete-continuous action space remains challenging. I"},"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":"2604.20328","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2026-04-22T08:22:23Z","cross_cats_sorted":[],"title_canon_sha256":"babc85231be9f3153e21fc0b7cd79da2bed899ffcbe63d3c8306193a988ca7ff","abstract_canon_sha256":"052aeb79afc3f64f03ca841590c441f74a38bde32704bb9880dd7ea95b2e35b4"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-29T01:14:31.379659Z","signature_b64":"uMl+viJeAwzOg/0Tt2lbD5SQOHl6zlNJUYcQc7YIEghJxySCkU6S/KyF7+huF1REq+dQF4lm54+Kjcth5Et1Ag==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"26bdc1634c940bc2e7f0e9d9c0d276d7e521aced7b4b7990269fc89d627a8187","last_reissued_at":"2026-06-29T01:14:31.379176Z","signature_status":"signed_v1","first_computed_at":"2026-06-29T01:14:31.379176Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"HyLaR: Hybrid Latent Reasoning with Decoupled Policy Optimization","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"Hybrid latent reasoning interleaves discrete text generation with continuous visual states and optimizes the combination through decoupled policy learning.","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Hao Zhang, Jinwen Luo, Shi-Zhe Chen, Tao Cheng, Yixin Qin, Zheng Wei","submitted_at":"2026-04-22T08:22:23Z","abstract_excerpt":"Chain-of-Thought (CoT) reasoning significantly elevates the complex problem-solving capabilities of multimodal large language models (MLLMs). However, adapting CoT to vision typically discretizes signals to fit LLM inputs, causing early semantic collapse and discarding fine-grained details. While external tools can mitigate this, they introduce a rigid bottleneck, confining reasoning to predefined operations. Although recent latent reasoning paradigms internalize visual states to overcome these limitations, optimizing the resulting hybrid discrete-continuous action space remains challenging. I"},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"HyLaR outperforms standard MLLMs and state-of-the-art latent reasoning approaches across fine-grained perception and general multimodal understanding benchmarks.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"The hybrid discrete-continuous action space can be effectively optimized via DePO with independent trust-region constraints and exact closed-form vMF KL regularizer without introducing instabilities or requiring extensive tuning.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"HyLaR with DePO enables effective RL in hybrid discrete-continuous spaces for multimodal models, outperforming prior MLLMs on perception and understanding benchmarks.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"Hybrid latent reasoning interleaves discrete text generation with continuous visual states and optimizes the combination through decoupled policy learning.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"7937486afadb009a9b8b6a15e42628e18f8badd3b1015be6769c4df06f16bacf"},"source":{"id":"2604.20328","kind":"arxiv","version":2},"verdict":{"id":"feef6044-d18c-4823-b458-e36c7fff8f9a","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-10T01:14:42.494900Z","strongest_claim":"HyLaR outperforms standard MLLMs and state-of-the-art latent reasoning approaches across fine-grained perception and general multimodal understanding benchmarks.","one_line_summary":"HyLaR with DePO enables effective RL in hybrid discrete-continuous spaces for multimodal models, outperforming prior MLLMs on perception and understanding benchmarks.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"The hybrid discrete-continuous action space can be effectively optimized via DePO with independent trust-region constraints and exact closed-form vMF KL regularizer without introducing instabilities or requiring extensive tuning.","pith_extraction_headline":"Hybrid latent reasoning interleaves discrete text generation with continuous visual states and optimizes the combination through decoupled policy learning."},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2604.20328/integrity.json","findings":[],"available":true,"detectors_run":[{"name":"ai_meta_artifact","ran_at":"2026-05-21T14:43:16.450368Z","status":"completed","version":"1.0.0","findings_count":0},{"name":"doi_compliance","ran_at":"2026-05-20T02:01:52.429681Z","status":"completed","version":"1.0.0","findings_count":0}],"snapshot_sha256":"ef30cdec3e955be5ea350d70b217a387a2e5dc6afad95f658985cdb89f63033c"},"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":"2604.20328","created_at":"2026-06-29T01:14:31.379232+00:00"},{"alias_kind":"arxiv_version","alias_value":"2604.20328v2","created_at":"2026-06-29T01:14:31.379232+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2604.20328","created_at":"2026-06-29T01:14:31.379232+00:00"},{"alias_kind":"pith_short_12","alias_value":"E264CY2MSQF4","created_at":"2026-06-29T01:14:31.379232+00:00"},{"alias_kind":"pith_short_16","alias_value":"E264CY2MSQF4FZ7Q","created_at":"2026-06-29T01:14:31.379232+00:00"},{"alias_kind":"pith_short_8","alias_value":"E264CY2M","created_at":"2026-06-29T01:14:31.379232+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2606.05769","citing_title":"Imagine Before You Predict: Interleaved Latent Visual Reasoning for Video Event Prediction","ref_index":83,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/E264CY2MSQF4FZ7Q5HM4BUTW27","json":"https://pith.science/pith/E264CY2MSQF4FZ7Q5HM4BUTW27.json","graph_json":"https://pith.science/api/pith-number/E264CY2MSQF4FZ7Q5HM4BUTW27/graph.json","events_json":"https://pith.science/api/pith-number/E264CY2MSQF4FZ7Q5HM4BUTW27/events.json","paper":"https://pith.science/paper/E264CY2M"},"agent_actions":{"view_html":"https://pith.science/pith/E264CY2MSQF4FZ7Q5HM4BUTW27","download_json":"https://pith.science/pith/E264CY2MSQF4FZ7Q5HM4BUTW27.json","view_paper":"https://pith.science/paper/E264CY2M","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2604.20328&json=true","fetch_graph":"https://pith.science/api/pith-number/E264CY2MSQF4FZ7Q5HM4BUTW27/graph.json","fetch_events":"https://pith.science/api/pith-number/E264CY2MSQF4FZ7Q5HM4BUTW27/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/E264CY2MSQF4FZ7Q5HM4BUTW27/action/timestamp_anchor","attest_storage":"https://pith.science/pith/E264CY2MSQF4FZ7Q5HM4BUTW27/action/storage_attestation","attest_author":"https://pith.science/pith/E264CY2MSQF4FZ7Q5HM4BUTW27/action/author_attestation","sign_citation":"https://pith.science/pith/E264CY2MSQF4FZ7Q5HM4BUTW27/action/citation_signature","submit_replication":"https://pith.science/pith/E264CY2MSQF4FZ7Q5HM4BUTW27/action/replication_record"}},"created_at":"2026-06-29T01:14:31.379232+00:00","updated_at":"2026-06-29T01:14:31.379232+00:00"}