{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2023:274J3NVRZT7I2ZM6NM75ONCSWF","short_pith_number":"pith:274J3NVR","schema_version":"1.0","canonical_sha256":"d7f89db6b1ccfe8d659e6b3fd73452b14fc3e1df92e0096f261ae4842de7bdef","source":{"kind":"arxiv","id":"2305.14453","version":2},"attestation_state":"computed","paper":{"title":"On Robustness of Finetuned Transformer-based NLP Models","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Manish Gupta, Mounika Marreddy, Pavan Kalyan Reddy Neerudu, Subba Reddy Oota, Venkateswara Rao Kagita","submitted_at":"2023-05-23T18:25:18Z","abstract_excerpt":"Transformer-based pretrained models like BERT, GPT-2 and T5 have been finetuned for a large number of natural language processing (NLP) tasks, and have been shown to be very effective. However, while finetuning, what changes across layers in these models with respect to pretrained checkpoints is under-studied. Further, how robust are these models to perturbations in input text? Does the robustness vary depending on the NLP task for which the models have been finetuned? While there exists some work on studying the robustness of BERT finetuned for a few NLP tasks, there is no rigorous study that"},"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":"2305.14453","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2023-05-23T18:25:18Z","cross_cats_sorted":[],"title_canon_sha256":"7625f6c3724a8c179f98ab319a603b87ff912137db59e092726189acb04e2436","abstract_canon_sha256":"8eb8b732c399a405a8a680464c6cc8c56818fb35a744d7bd78063c392eabe617"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T07:10:24.923554Z","signature_b64":"FFDmpLO+hA2u4ueX3sKU+ZvN3FNCYzAejFwPFtMQu/T6mJm8y4J53YMZFV8W9jq9h16TNrYx8HXhE5cCxecqDA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"d7f89db6b1ccfe8d659e6b3fd73452b14fc3e1df92e0096f261ae4842de7bdef","last_reissued_at":"2026-07-05T07:10:24.923041Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T07:10:24.923041Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"On Robustness of Finetuned Transformer-based NLP Models","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Manish Gupta, Mounika Marreddy, Pavan Kalyan Reddy Neerudu, Subba Reddy Oota, Venkateswara Rao Kagita","submitted_at":"2023-05-23T18:25:18Z","abstract_excerpt":"Transformer-based pretrained models like BERT, GPT-2 and T5 have been finetuned for a large number of natural language processing (NLP) tasks, and have been shown to be very effective. However, while finetuning, what changes across layers in these models with respect to pretrained checkpoints is under-studied. Further, how robust are these models to perturbations in input text? Does the robustness vary depending on the NLP task for which the models have been finetuned? While there exists some work on studying the robustness of BERT finetuned for a few NLP tasks, there is no rigorous study that"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2305.14453","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/2305.14453/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":"2305.14453","created_at":"2026-07-05T07:10:24.923097+00:00"},{"alias_kind":"arxiv_version","alias_value":"2305.14453v2","created_at":"2026-07-05T07:10:24.923097+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2305.14453","created_at":"2026-07-05T07:10:24.923097+00:00"},{"alias_kind":"pith_short_12","alias_value":"274J3NVRZT7I","created_at":"2026-07-05T07:10:24.923097+00:00"},{"alias_kind":"pith_short_16","alias_value":"274J3NVRZT7I2ZM6","created_at":"2026-07-05T07:10:24.923097+00:00"},{"alias_kind":"pith_short_8","alias_value":"274J3NVR","created_at":"2026-07-05T07:10:24.923097+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2508.18306","citing_title":"SALMAN: Stability Analysis of Language Models Through the Maps Between Graph-based Manifolds","ref_index":26,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/274J3NVRZT7I2ZM6NM75ONCSWF","json":"https://pith.science/pith/274J3NVRZT7I2ZM6NM75ONCSWF.json","graph_json":"https://pith.science/api/pith-number/274J3NVRZT7I2ZM6NM75ONCSWF/graph.json","events_json":"https://pith.science/api/pith-number/274J3NVRZT7I2ZM6NM75ONCSWF/events.json","paper":"https://pith.science/paper/274J3NVR"},"agent_actions":{"view_html":"https://pith.science/pith/274J3NVRZT7I2ZM6NM75ONCSWF","download_json":"https://pith.science/pith/274J3NVRZT7I2ZM6NM75ONCSWF.json","view_paper":"https://pith.science/paper/274J3NVR","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2305.14453&json=true","fetch_graph":"https://pith.science/api/pith-number/274J3NVRZT7I2ZM6NM75ONCSWF/graph.json","fetch_events":"https://pith.science/api/pith-number/274J3NVRZT7I2ZM6NM75ONCSWF/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/274J3NVRZT7I2ZM6NM75ONCSWF/action/timestamp_anchor","attest_storage":"https://pith.science/pith/274J3NVRZT7I2ZM6NM75ONCSWF/action/storage_attestation","attest_author":"https://pith.science/pith/274J3NVRZT7I2ZM6NM75ONCSWF/action/author_attestation","sign_citation":"https://pith.science/pith/274J3NVRZT7I2ZM6NM75ONCSWF/action/citation_signature","submit_replication":"https://pith.science/pith/274J3NVRZT7I2ZM6NM75ONCSWF/action/replication_record"}},"created_at":"2026-07-05T07:10:24.923097+00:00","updated_at":"2026-07-05T07:10:24.923097+00:00"}