{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2020:K57RXHY47QIKLA32ZNKSQO6PLS","short_pith_number":"pith:K57RXHY4","schema_version":"1.0","canonical_sha256":"577f1b9f1cfc10a5837acb55283bcf5ca4d599192899725e564c5d930cf34126","source":{"kind":"arxiv","id":"2004.03685","version":3},"attestation_state":"computed","paper":{"title":"Towards Faithfully Interpretable NLP Systems: How should we define and evaluate faithfulness?","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.CL","authors_text":"Alon Jacovi, Yoav Goldberg","submitted_at":"2020-04-07T20:15:28Z","abstract_excerpt":"With the growing popularity of deep-learning based NLP models, comes a need for interpretable systems. But what is interpretability, and what constitutes a high-quality interpretation? In this opinion piece we reflect on the current state of interpretability evaluation research. We call for more clearly differentiating between different desired criteria an interpretation should satisfy, and focus on the faithfulness criteria. We survey the literature with respect to faithfulness evaluation, and arrange the current approaches around three assumptions, providing an explicit form to how faithfuln"},"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":"2004.03685","kind":"arxiv","version":3},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2020-04-07T20:15:28Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"c6399a37adc899a42b6a126ec152222f9c7563de608861aabef96094c7f118ca","abstract_canon_sha256":"be399f7fc583b0ceedac2f19b8dec0177529f9282d3d46811a466fb364d99e4e"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T00:58:52.300493Z","signature_b64":"B1T4yU0JQyyiAeFwLeiP5MYfBixRNZM90EoqfXO+dXZwRJZCSiOevjpqLsGZbbxhpu1oRSOFynCJAKoQ1WugDg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"577f1b9f1cfc10a5837acb55283bcf5ca4d599192899725e564c5d930cf34126","last_reissued_at":"2026-07-05T00:58:52.300013Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T00:58:52.300013Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Towards Faithfully Interpretable NLP Systems: How should we define and evaluate faithfulness?","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.CL","authors_text":"Alon Jacovi, Yoav Goldberg","submitted_at":"2020-04-07T20:15:28Z","abstract_excerpt":"With the growing popularity of deep-learning based NLP models, comes a need for interpretable systems. But what is interpretability, and what constitutes a high-quality interpretation? In this opinion piece we reflect on the current state of interpretability evaluation research. We call for more clearly differentiating between different desired criteria an interpretation should satisfy, and focus on the faithfulness criteria. We survey the literature with respect to faithfulness evaluation, and arrange the current approaches around three assumptions, providing an explicit form to how faithfuln"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2004.03685","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/2004.03685/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":"2004.03685","created_at":"2026-07-05T00:58:52.300061+00:00"},{"alias_kind":"arxiv_version","alias_value":"2004.03685v3","created_at":"2026-07-05T00:58:52.300061+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2004.03685","created_at":"2026-07-05T00:58:52.300061+00:00"},{"alias_kind":"pith_short_12","alias_value":"K57RXHY47QIK","created_at":"2026-07-05T00:58:52.300061+00:00"},{"alias_kind":"pith_short_16","alias_value":"K57RXHY47QIKLA32","created_at":"2026-07-05T00:58:52.300061+00:00"},{"alias_kind":"pith_short_8","alias_value":"K57RXHY4","created_at":"2026-07-05T00:58:52.300061+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":11,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.27274","citing_title":"BetXplain: An Explanation-Annotated Dataset for Detecting Manipulative Betting Advertisements on Social Media","ref_index":162,"is_internal_anchor":false},{"citing_arxiv_id":"2606.04454","citing_title":"Stepwise Reasoning Enhancement for LLMs via External Subgraph Generation","ref_index":3,"is_internal_anchor":false},{"citing_arxiv_id":"2605.27879","citing_title":"Towards Faithful Agentic XAI: A Verification Method and an Open-World Benchmark for Better Model Faithfulness","ref_index":2,"is_internal_anchor":false},{"citing_arxiv_id":"2503.16771","citing_title":"Enabling Global, Human-Centered Explanations for LLMs:From Tokens to Interpretable Code and Test Generation","ref_index":25,"is_internal_anchor":false},{"citing_arxiv_id":"2605.21849","citing_title":"Geometry-Adaptive Explainer for Faithful Dictionary-Based Interpretability under Distribution Shift","ref_index":9,"is_internal_anchor":false},{"citing_arxiv_id":"2503.11926","citing_title":"Monitoring Reasoning Models for Misbehavior and the Risks of Promoting Obfuscation","ref_index":27,"is_internal_anchor":false},{"citing_arxiv_id":"2605.16117","citing_title":"SGR: A Stepwise Reasoning Framework for LLMs with External Subgraph Generation","ref_index":3,"is_internal_anchor":false},{"citing_arxiv_id":"2605.17562","citing_title":"Beyond Accuracy: Robustness, Interpretability and Expressiveness of EEG Foundation Models","ref_index":46,"is_internal_anchor":false},{"citing_arxiv_id":"2605.18660","citing_title":"Evaluating Multi-turn Human-AI Interaction","ref_index":68,"is_internal_anchor":false},{"citing_arxiv_id":"2605.12809","citing_title":"Correcting Influence: Unboxing LLM Outputs with Orthogonal Latent Spaces","ref_index":174,"is_internal_anchor":false},{"citing_arxiv_id":"2604.16158","citing_title":"AtManRL: Towards Faithful Reasoning via Differentiable Attention Saliency","ref_index":8,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/K57RXHY47QIKLA32ZNKSQO6PLS","json":"https://pith.science/pith/K57RXHY47QIKLA32ZNKSQO6PLS.json","graph_json":"https://pith.science/api/pith-number/K57RXHY47QIKLA32ZNKSQO6PLS/graph.json","events_json":"https://pith.science/api/pith-number/K57RXHY47QIKLA32ZNKSQO6PLS/events.json","paper":"https://pith.science/paper/K57RXHY4"},"agent_actions":{"view_html":"https://pith.science/pith/K57RXHY47QIKLA32ZNKSQO6PLS","download_json":"https://pith.science/pith/K57RXHY47QIKLA32ZNKSQO6PLS.json","view_paper":"https://pith.science/paper/K57RXHY4","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2004.03685&json=true","fetch_graph":"https://pith.science/api/pith-number/K57RXHY47QIKLA32ZNKSQO6PLS/graph.json","fetch_events":"https://pith.science/api/pith-number/K57RXHY47QIKLA32ZNKSQO6PLS/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/K57RXHY47QIKLA32ZNKSQO6PLS/action/timestamp_anchor","attest_storage":"https://pith.science/pith/K57RXHY47QIKLA32ZNKSQO6PLS/action/storage_attestation","attest_author":"https://pith.science/pith/K57RXHY47QIKLA32ZNKSQO6PLS/action/author_attestation","sign_citation":"https://pith.science/pith/K57RXHY47QIKLA32ZNKSQO6PLS/action/citation_signature","submit_replication":"https://pith.science/pith/K57RXHY47QIKLA32ZNKSQO6PLS/action/replication_record"}},"created_at":"2026-07-05T00:58:52.300061+00:00","updated_at":"2026-07-05T00:58:52.300061+00:00"}