{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2023:UM2YPTSV2PVSU7NKV72UNM4CLS","short_pith_number":"pith:UM2YPTSV","schema_version":"1.0","canonical_sha256":"a33587ce55d3eb2a7daaaff546b3825cb16c66cf09a76e54a6cea50961a89cbf","source":{"kind":"arxiv","id":"2310.06117","version":2},"attestation_state":"computed","paper":{"title":"Take a Step Back: Evoking Reasoning via Abstraction in Large Language Models","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.CL"],"primary_cat":"cs.LG","authors_text":"Denny Zhou, Ed H. Chi, Heng-Tze Cheng, Huaixiu Steven Zheng, Quoc V Le, Swaroop Mishra, Xinyun Chen","submitted_at":"2023-10-09T19:48:55Z","abstract_excerpt":"We present Step-Back Prompting, a simple prompting technique that enables LLMs to do abstractions to derive high-level concepts and first principles from instances containing specific details. Using the concepts and principles to guide reasoning, LLMs significantly improve their abilities in following a correct reasoning path towards the solution. We conduct experiments of Step-Back Prompting with PaLM-2L, GPT-4 and Llama2-70B models, and observe substantial performance gains on various challenging reasoning-intensive tasks including STEM, Knowledge QA, and Multi-Hop Reasoning. For instance, S"},"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":"2310.06117","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2023-10-09T19:48:55Z","cross_cats_sorted":["cs.AI","cs.CL"],"title_canon_sha256":"179788da0b4fa33bf1aa706ea4838f9c81d47d976375b913b29e1fb6ba19644e","abstract_canon_sha256":"fbe7a5977f44d1d60ef0d1f28030c7b32a67a19cd5990c9ec7dc711afb5191bd"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T07:54:51.774877Z","signature_b64":"EZLHMLCubBIvQhVGCnCE5bBxEng1MA9EF3cpDdEEafGg4bZYKX6XrTU5nhGU0wDBNx/YwRoLs28eTXwO/uH3CA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"a33587ce55d3eb2a7daaaff546b3825cb16c66cf09a76e54a6cea50961a89cbf","last_reissued_at":"2026-07-05T07:54:51.774430Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T07:54:51.774430Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Take a Step Back: Evoking Reasoning via Abstraction in Large Language Models","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.CL"],"primary_cat":"cs.LG","authors_text":"Denny Zhou, Ed H. Chi, Heng-Tze Cheng, Huaixiu Steven Zheng, Quoc V Le, Swaroop Mishra, Xinyun Chen","submitted_at":"2023-10-09T19:48:55Z","abstract_excerpt":"We present Step-Back Prompting, a simple prompting technique that enables LLMs to do abstractions to derive high-level concepts and first principles from instances containing specific details. Using the concepts and principles to guide reasoning, LLMs significantly improve their abilities in following a correct reasoning path towards the solution. We conduct experiments of Step-Back Prompting with PaLM-2L, GPT-4 and Llama2-70B models, and observe substantial performance gains on various challenging reasoning-intensive tasks including STEM, Knowledge QA, and Multi-Hop Reasoning. For instance, S"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2310.06117","kind":"arxiv","version":2},"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/2310.06117/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":"2310.06117","created_at":"2026-07-05T07:54:51.774493+00:00"},{"alias_kind":"arxiv_version","alias_value":"2310.06117v2","created_at":"2026-07-05T07:54:51.774493+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2310.06117","created_at":"2026-07-05T07:54:51.774493+00:00"},{"alias_kind":"pith_short_12","alias_value":"UM2YPTSV2PVS","created_at":"2026-07-05T07:54:51.774493+00:00"},{"alias_kind":"pith_short_16","alias_value":"UM2YPTSV2PVSU7NK","created_at":"2026-07-05T07:54:51.774493+00:00"},{"alias_kind":"pith_short_8","alias_value":"UM2YPTSV","created_at":"2026-07-05T07:54:51.774493+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":14,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.22385","citing_title":"MetaPS: Adaptive Programmatic Strategy Selection for Market Agents","ref_index":24,"is_internal_anchor":false},{"citing_arxiv_id":"2606.01441","citing_title":"Dive into Ambiguity: A*-Inspired Multi-Agents Commonsense Obfuscation Attack on LLM Prompts","ref_index":19,"is_internal_anchor":false},{"citing_arxiv_id":"2606.01240","citing_title":"Efficient RAG with Intent-Aware Retrieval and Semantics-Preserving Chunking","ref_index":17,"is_internal_anchor":false},{"citing_arxiv_id":"2312.10997","citing_title":"Retrieval-Augmented Generation for Large Language Models: A Survey","ref_index":10,"is_internal_anchor":false},{"citing_arxiv_id":"2404.10981","citing_title":"A Survey on Retrieval-Augmented Text Generation for Large Language Models","ref_index":171,"is_internal_anchor":false},{"citing_arxiv_id":"2409.10102","citing_title":"Trustworthiness in Retrieval-Augmented Generation Systems: A Survey","ref_index":34,"is_internal_anchor":false},{"citing_arxiv_id":"2509.22202","citing_title":"Library Hallucinations in LLM-Generated Code: A Risk Analysis Grounded in Developer Queries","ref_index":77,"is_internal_anchor":false},{"citing_arxiv_id":"2605.17187","citing_title":"PluRule: A Benchmark for Moderating Pluralistic Communities on Social Media","ref_index":99,"is_internal_anchor":false},{"citing_arxiv_id":"2402.19473","citing_title":"Retrieval-Augmented Generation for AI-Generated Content: A Survey","ref_index":167,"is_internal_anchor":false},{"citing_arxiv_id":"2408.08435","citing_title":"Automated Design of Agentic Systems","ref_index":237,"is_internal_anchor":false},{"citing_arxiv_id":"2605.11344","citing_title":"Making Abstraction Concrete: A Design Space and Interaction Model of Abstraction in Interactive Systems","ref_index":202,"is_internal_anchor":false},{"citing_arxiv_id":"2402.07927","citing_title":"A Systematic Survey of Prompt Engineering in Large Language Models: Techniques and Applications","ref_index":23,"is_internal_anchor":false},{"citing_arxiv_id":"2605.03950","citing_title":"UnAC: Adaptive Visual Prompting with Abstraction and Stepwise Checking for Complex Multimodal Reasoning","ref_index":19,"is_internal_anchor":false},{"citing_arxiv_id":"2605.00676","citing_title":"Living Databases: A Unified Model for Continuous Schema Evolution, Versioning, and Transformations","ref_index":81,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/UM2YPTSV2PVSU7NKV72UNM4CLS","json":"https://pith.science/pith/UM2YPTSV2PVSU7NKV72UNM4CLS.json","graph_json":"https://pith.science/api/pith-number/UM2YPTSV2PVSU7NKV72UNM4CLS/graph.json","events_json":"https://pith.science/api/pith-number/UM2YPTSV2PVSU7NKV72UNM4CLS/events.json","paper":"https://pith.science/paper/UM2YPTSV"},"agent_actions":{"view_html":"https://pith.science/pith/UM2YPTSV2PVSU7NKV72UNM4CLS","download_json":"https://pith.science/pith/UM2YPTSV2PVSU7NKV72UNM4CLS.json","view_paper":"https://pith.science/paper/UM2YPTSV","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2310.06117&json=true","fetch_graph":"https://pith.science/api/pith-number/UM2YPTSV2PVSU7NKV72UNM4CLS/graph.json","fetch_events":"https://pith.science/api/pith-number/UM2YPTSV2PVSU7NKV72UNM4CLS/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/UM2YPTSV2PVSU7NKV72UNM4CLS/action/timestamp_anchor","attest_storage":"https://pith.science/pith/UM2YPTSV2PVSU7NKV72UNM4CLS/action/storage_attestation","attest_author":"https://pith.science/pith/UM2YPTSV2PVSU7NKV72UNM4CLS/action/author_attestation","sign_citation":"https://pith.science/pith/UM2YPTSV2PVSU7NKV72UNM4CLS/action/citation_signature","submit_replication":"https://pith.science/pith/UM2YPTSV2PVSU7NKV72UNM4CLS/action/replication_record"}},"created_at":"2026-07-05T07:54:51.774493+00:00","updated_at":"2026-07-05T07:54:51.774493+00:00"}