{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:RFGFZQMGMWAIDWBCRN3FECN2IA","short_pith_number":"pith:RFGFZQMG","schema_version":"1.0","canonical_sha256":"894c5cc186658081d8228b765209ba40383116727725da4f436cd61b20866a25","source":{"kind":"arxiv","id":"2505.18962","version":3},"attestation_state":"computed","paper":{"title":"System-1.5 Reasoning: Traversal in Language and Latent Spaces with Dynamic Shortcuts","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Bang Liu, Suyuchen Wang, Xiaoqiang Wang, Yun Zhu","submitted_at":"2025-05-25T03:35:49Z","abstract_excerpt":"Chain-of-thought (CoT) reasoning enables large language models (LLMs) to move beyond fast System-1 responses and engage in deliberative System-2 reasoning. However, this comes at the cost of significant inefficiency due to verbose intermediate output. Recent latent-space reasoning methods improve efficiency by operating on hidden states without decoding into language, yet they treat all steps uniformly, failing to distinguish critical deductions from auxiliary steps and resulting in suboptimal use of computational resources. In this paper, we propose System-1.5 Reasoning, an adaptive reasoning"},"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":"2505.18962","kind":"arxiv","version":3},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2025-05-25T03:35:49Z","cross_cats_sorted":[],"title_canon_sha256":"ce59fad08d60db0c4eb40f13e47ab71b51947bedac421eaab21883dbed74fe04","abstract_canon_sha256":"cd258d57666c0e18e8ece85caddfc2d279d5f657f8cd8b4cb4fb430b512d3e56"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:13:26.683033Z","signature_b64":"P1zKPd5sAId/7QsSCg47nBMac7OMRqjRwoAkxgELkrD/fIz/qS+Hc2FhVc1wLnAlFQU+94K8FL2mk/9eJNQ7Cg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"894c5cc186658081d8228b765209ba40383116727725da4f436cd61b20866a25","last_reissued_at":"2026-07-05T11:13:26.682595Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:13:26.682595Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"System-1.5 Reasoning: Traversal in Language and Latent Spaces with Dynamic Shortcuts","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Bang Liu, Suyuchen Wang, Xiaoqiang Wang, Yun Zhu","submitted_at":"2025-05-25T03:35:49Z","abstract_excerpt":"Chain-of-thought (CoT) reasoning enables large language models (LLMs) to move beyond fast System-1 responses and engage in deliberative System-2 reasoning. However, this comes at the cost of significant inefficiency due to verbose intermediate output. Recent latent-space reasoning methods improve efficiency by operating on hidden states without decoding into language, yet they treat all steps uniformly, failing to distinguish critical deductions from auxiliary steps and resulting in suboptimal use of computational resources. In this paper, we propose System-1.5 Reasoning, an adaptive reasoning"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2505.18962","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/2505.18962/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":"2505.18962","created_at":"2026-07-05T11:13:26.682652+00:00"},{"alias_kind":"arxiv_version","alias_value":"2505.18962v3","created_at":"2026-07-05T11:13:26.682652+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2505.18962","created_at":"2026-07-05T11:13:26.682652+00:00"},{"alias_kind":"pith_short_12","alias_value":"RFGFZQMGMWAI","created_at":"2026-07-05T11:13:26.682652+00:00"},{"alias_kind":"pith_short_16","alias_value":"RFGFZQMGMWAIDWBC","created_at":"2026-07-05T11:13:26.682652+00:00"},{"alias_kind":"pith_short_8","alias_value":"RFGFZQMG","created_at":"2026-07-05T11:13:26.682652+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":6,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2504.01990","citing_title":"Advances and Challenges in Foundation Agents: From Brain-Inspired Intelligence to Evolutionary, Collaborative, and Safe Systems","ref_index":196,"is_internal_anchor":false},{"citing_arxiv_id":"2605.22675","citing_title":"Self-Policy Distillation via Capability-Selective Subspace Projection","ref_index":23,"is_internal_anchor":false},{"citing_arxiv_id":"2605.17329","citing_title":"LPG: Balancing Efficiency and Policy Reasoning in Latent Policy Guardrails","ref_index":32,"is_internal_anchor":false},{"citing_arxiv_id":"2604.03679","citing_title":"LightThinker++: From Reasoning Compression to Memory Management","ref_index":78,"is_internal_anchor":false},{"citing_arxiv_id":"2604.17866","citing_title":"Latent Abstraction for Retrieval-Augmented Generation","ref_index":40,"is_internal_anchor":false},{"citing_arxiv_id":"2604.08299","citing_title":"SeLaR: Selective Latent Reasoning in Large Language Models","ref_index":42,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/RFGFZQMGMWAIDWBCRN3FECN2IA","json":"https://pith.science/pith/RFGFZQMGMWAIDWBCRN3FECN2IA.json","graph_json":"https://pith.science/api/pith-number/RFGFZQMGMWAIDWBCRN3FECN2IA/graph.json","events_json":"https://pith.science/api/pith-number/RFGFZQMGMWAIDWBCRN3FECN2IA/events.json","paper":"https://pith.science/paper/RFGFZQMG"},"agent_actions":{"view_html":"https://pith.science/pith/RFGFZQMGMWAIDWBCRN3FECN2IA","download_json":"https://pith.science/pith/RFGFZQMGMWAIDWBCRN3FECN2IA.json","view_paper":"https://pith.science/paper/RFGFZQMG","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2505.18962&json=true","fetch_graph":"https://pith.science/api/pith-number/RFGFZQMGMWAIDWBCRN3FECN2IA/graph.json","fetch_events":"https://pith.science/api/pith-number/RFGFZQMGMWAIDWBCRN3FECN2IA/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/RFGFZQMGMWAIDWBCRN3FECN2IA/action/timestamp_anchor","attest_storage":"https://pith.science/pith/RFGFZQMGMWAIDWBCRN3FECN2IA/action/storage_attestation","attest_author":"https://pith.science/pith/RFGFZQMGMWAIDWBCRN3FECN2IA/action/author_attestation","sign_citation":"https://pith.science/pith/RFGFZQMGMWAIDWBCRN3FECN2IA/action/citation_signature","submit_replication":"https://pith.science/pith/RFGFZQMGMWAIDWBCRN3FECN2IA/action/replication_record"}},"created_at":"2026-07-05T11:13:26.682652+00:00","updated_at":"2026-07-05T11:13:26.682652+00:00"}