{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2025:JJSGN4OBPHV5IHDCXKEMRGUXOA","short_pith_number":"pith:JJSGN4OB","canonical_record":{"source":{"id":"2503.13208","kind":"arxiv","version":3},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-03-17T14:20:48Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"149af4fe45e3f0ab14a0bd9ecca35dad89c4441fd0d3e02327b28fc8bdefb147","abstract_canon_sha256":"74520bd1ae8d59daec307bae5ecf9be9428ab6f65880fee8be9de1525df69849"},"schema_version":"1.0"},"canonical_sha256":"4a6466f1c179ebd41c62ba88c89a97702d367ffd8cac2f7acf2942a2a8f682df","source":{"kind":"arxiv","id":"2503.13208","version":3},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2503.13208","created_at":"2026-07-05T10:48:27Z"},{"alias_kind":"arxiv_version","alias_value":"2503.13208v3","created_at":"2026-07-05T10:48:27Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2503.13208","created_at":"2026-07-05T10:48:27Z"},{"alias_kind":"pith_short_12","alias_value":"JJSGN4OBPHV5","created_at":"2026-07-05T10:48:27Z"},{"alias_kind":"pith_short_16","alias_value":"JJSGN4OBPHV5IHDC","created_at":"2026-07-05T10:48:27Z"},{"alias_kind":"pith_short_8","alias_value":"JJSGN4OB","created_at":"2026-07-05T10:48:27Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2025:JJSGN4OBPHV5IHDCXKEMRGUXOA","target":"record","payload":{"canonical_record":{"source":{"id":"2503.13208","kind":"arxiv","version":3},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-03-17T14:20:48Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"149af4fe45e3f0ab14a0bd9ecca35dad89c4441fd0d3e02327b28fc8bdefb147","abstract_canon_sha256":"74520bd1ae8d59daec307bae5ecf9be9428ab6f65880fee8be9de1525df69849"},"schema_version":"1.0"},"canonical_sha256":"4a6466f1c179ebd41c62ba88c89a97702d367ffd8cac2f7acf2942a2a8f682df","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:48:27.200736Z","signature_b64":"a1943AlLZquUVzjtSNe3NqDpvhZscwEijR3e0Or2aoXazRI5oxEmEErLvV2qZxNxUsiDaaQGdAYtOcuclx4WBA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"4a6466f1c179ebd41c62ba88c89a97702d367ffd8cac2f7acf2942a2a8f682df","last_reissued_at":"2026-07-05T10:48:27.200265Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:48:27.200265Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2503.13208","source_version":3,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T10:48:27Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"Ah2bCk3gmxfhVaNeR4qYaUlx1wBpUb9U8OAEAYo7M5Iwh3V0n/M4V92fRHyQqQbmhiGcNUPIXt9Hreol9gp+BA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-20T02:22:21.755389Z"},"content_sha256":"3910ebebd90c2d00be9c121acbaa5d9ed9438f3d669f492982b1ba6697d956e0","schema_version":"1.0","event_id":"sha256:3910ebebd90c2d00be9c121acbaa5d9ed9438f3d669f492982b1ba6697d956e0"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2025:JJSGN4OBPHV5IHDCXKEMRGUXOA","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Improving Complex Reasoning with Dynamic Prompt Corruption: A soft prompt Optimization Approach","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Chen Shen, Chenxi Huang, Ge Teng, Jieping Ye, Liang Xie, Sinan Fan, Wenxiao Wang, Xiaofei He, Xiaofeng Zhang, Xiaosong Yuan","submitted_at":"2025-03-17T14:20:48Z","abstract_excerpt":"Prompt-tuning (PT) for large language models (LLMs) can facilitate the performance on various conventional NLP tasks with significantly fewer trainable parameters. However, our investigation reveals that PT provides limited improvement and may even degrade the primitive performance of LLMs on complex reasoning tasks. Such a phenomenon suggests that soft prompts can positively impact certain instances while negatively affecting others, particularly during the later phases of reasoning. To address these challenges, We first identify an information accumulation within the soft prompts. Through de"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2503.13208","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/2503.13208/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":"2026-07-05T10:48:27Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"rmkapac1RjYfHHRwLIdt3VfnpXJk1r0UjRyehYs9TkR5gD1k1DAuMyOx0q5ZMvwLSRqLq6AY++sL74MyxizYCA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-20T02:22:21.756062Z"},"content_sha256":"e31f86d5929db5c36558035b8b8127e756c6ee20d3585a19ad2c2988f00b68e6","schema_version":"1.0","event_id":"sha256:e31f86d5929db5c36558035b8b8127e756c6ee20d3585a19ad2c2988f00b68e6"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/JJSGN4OBPHV5IHDCXKEMRGUXOA/bundle.json","state_url":"https://pith.science/pith/JJSGN4OBPHV5IHDCXKEMRGUXOA/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/JJSGN4OBPHV5IHDCXKEMRGUXOA/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":"2026-08-20T02:22:21Z","links":{"resolver":"https://pith.science/pith/JJSGN4OBPHV5IHDCXKEMRGUXOA","bundle":"https://pith.science/pith/JJSGN4OBPHV5IHDCXKEMRGUXOA/bundle.json","state":"https://pith.science/pith/JJSGN4OBPHV5IHDCXKEMRGUXOA/state.json","well_known_bundle":"https://pith.science/.well-known/pith/JJSGN4OBPHV5IHDCXKEMRGUXOA/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2025:JJSGN4OBPHV5IHDCXKEMRGUXOA","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":"74520bd1ae8d59daec307bae5ecf9be9428ab6f65880fee8be9de1525df69849","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-03-17T14:20:48Z","title_canon_sha256":"149af4fe45e3f0ab14a0bd9ecca35dad89c4441fd0d3e02327b28fc8bdefb147"},"schema_version":"1.0","source":{"id":"2503.13208","kind":"arxiv","version":3}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2503.13208","created_at":"2026-07-05T10:48:27Z"},{"alias_kind":"arxiv_version","alias_value":"2503.13208v3","created_at":"2026-07-05T10:48:27Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2503.13208","created_at":"2026-07-05T10:48:27Z"},{"alias_kind":"pith_short_12","alias_value":"JJSGN4OBPHV5","created_at":"2026-07-05T10:48:27Z"},{"alias_kind":"pith_short_16","alias_value":"JJSGN4OBPHV5IHDC","created_at":"2026-07-05T10:48:27Z"},{"alias_kind":"pith_short_8","alias_value":"JJSGN4OB","created_at":"2026-07-05T10:48:27Z"}],"graph_snapshots":[{"event_id":"sha256:e31f86d5929db5c36558035b8b8127e756c6ee20d3585a19ad2c2988f00b68e6","target":"graph","created_at":"2026-07-05T10:48:27Z","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/2503.13208/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Prompt-tuning (PT) for large language models (LLMs) can facilitate the performance on various conventional NLP tasks with significantly fewer trainable parameters. However, our investigation reveals that PT provides limited improvement and may even degrade the primitive performance of LLMs on complex reasoning tasks. Such a phenomenon suggests that soft prompts can positively impact certain instances while negatively affecting others, particularly during the later phases of reasoning. To address these challenges, We first identify an information accumulation within the soft prompts. Through de","authors_text":"Chen Shen, Chenxi Huang, Ge Teng, Jieping Ye, Liang Xie, Sinan Fan, Wenxiao Wang, Xiaofei He, Xiaofeng Zhang, Xiaosong Yuan","cross_cats":["cs.AI"],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-03-17T14:20:48Z","title":"Improving Complex Reasoning with Dynamic Prompt Corruption: A soft prompt Optimization Approach"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2503.13208","kind":"arxiv","version":3},"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":"sha256:3910ebebd90c2d00be9c121acbaa5d9ed9438f3d669f492982b1ba6697d956e0","target":"record","created_at":"2026-07-05T10:48:27Z","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":"74520bd1ae8d59daec307bae5ecf9be9428ab6f65880fee8be9de1525df69849","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-03-17T14:20:48Z","title_canon_sha256":"149af4fe45e3f0ab14a0bd9ecca35dad89c4441fd0d3e02327b28fc8bdefb147"},"schema_version":"1.0","source":{"id":"2503.13208","kind":"arxiv","version":3}},"canonical_sha256":"4a6466f1c179ebd41c62ba88c89a97702d367ffd8cac2f7acf2942a2a8f682df","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"4a6466f1c179ebd41c62ba88c89a97702d367ffd8cac2f7acf2942a2a8f682df","first_computed_at":"2026-07-05T10:48:27.200265Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T10:48:27.200265Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"a1943AlLZquUVzjtSNe3NqDpvhZscwEijR3e0Or2aoXazRI5oxEmEErLvV2qZxNxUsiDaaQGdAYtOcuclx4WBA==","signature_status":"signed_v1","signed_at":"2026-07-05T10:48:27.200736Z","signed_message":"canonical_sha256_bytes"},"source_id":"2503.13208","source_kind":"arxiv","source_version":3}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:3910ebebd90c2d00be9c121acbaa5d9ed9438f3d669f492982b1ba6697d956e0","sha256:e31f86d5929db5c36558035b8b8127e756c6ee20d3585a19ad2c2988f00b68e6"],"state_sha256":"275a510160583e6480ff5e4d95ebd30764da0dc79ba1b5c82c7a3eed716017e6"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"gFC8YHToMXcTDkQn3hsnzLcOLVlaDOVK8SautcrkFOvmSreuS1ZqB4RLi88vvFpPXpncAI7W9Ij0fKORSiKsDw==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-20T02:22:21.760354Z","bundle_sha256":"4e40818663e97f42847fd438b173727062e0a8f2a91e43ca90327e24d2143f94"}}