{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:RNYLAY3ZDL375XY2DUFK5EXNPQ","short_pith_number":"pith:RNYLAY3Z","schema_version":"1.0","canonical_sha256":"8b70b063791af7fedf1a1d0aae92ed7c0fa9e9b792e0dbbba028ce2c763a6e46","source":{"kind":"arxiv","id":"2405.16714","version":1},"attestation_state":"computed","paper":{"title":"Crafting Interpretable Embeddings by Asking LLMs Questions","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.LG","q-bio.NC"],"primary_cat":"cs.CL","authors_text":"Alexander G. Huth, Chandan Singh, Ion Stoica, Jianfeng Gao, John X. Morris, Richard Antonello, Vinamra Benara","submitted_at":"2024-05-26T22:30:29Z","abstract_excerpt":"Large language models (LLMs) have rapidly improved text embeddings for a growing array of natural-language processing tasks. However, their opaqueness and proliferation into scientific domains such as neuroscience have created a growing need for interpretability. Here, we ask whether we can obtain interpretable embeddings through LLM prompting. We introduce question-answering embeddings (QA-Emb), embeddings where each feature represents an answer to a yes/no question asked to an LLM. Training QA-Emb reduces to selecting a set of underlying questions rather than learning model weights.\n  We use"},"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":"2405.16714","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-05-26T22:30:29Z","cross_cats_sorted":["cs.AI","cs.LG","q-bio.NC"],"title_canon_sha256":"323626e4176c62bb97c7233ae1024919c32ed25d5f1da5e11a5705178d23ead3","abstract_canon_sha256":"30e51e10b4d8c39c54135e5aadcf53b3b840343ec6bf888a8f6bbed1b9585999"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T08:23:34.415187Z","signature_b64":"F+yXLJAhg8H0DjdBcwtk3NqSlt1o3Ux1utEXOrVSR47tj8ZYOl2vqL3NlZIaKSDR5CaJfC0tYRcAPsL/QPYuDg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"8b70b063791af7fedf1a1d0aae92ed7c0fa9e9b792e0dbbba028ce2c763a6e46","last_reissued_at":"2026-07-05T08:23:34.414708Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T08:23:34.414708Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Crafting Interpretable Embeddings by Asking LLMs Questions","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.LG","q-bio.NC"],"primary_cat":"cs.CL","authors_text":"Alexander G. Huth, Chandan Singh, Ion Stoica, Jianfeng Gao, John X. Morris, Richard Antonello, Vinamra Benara","submitted_at":"2024-05-26T22:30:29Z","abstract_excerpt":"Large language models (LLMs) have rapidly improved text embeddings for a growing array of natural-language processing tasks. However, their opaqueness and proliferation into scientific domains such as neuroscience have created a growing need for interpretability. Here, we ask whether we can obtain interpretable embeddings through LLM prompting. We introduce question-answering embeddings (QA-Emb), embeddings where each feature represents an answer to a yes/no question asked to an LLM. Training QA-Emb reduces to selecting a set of underlying questions rather than learning model weights.\n  We use"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2405.16714","kind":"arxiv","version":1},"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/2405.16714/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":"2405.16714","created_at":"2026-07-05T08:23:34.414766+00:00"},{"alias_kind":"arxiv_version","alias_value":"2405.16714v1","created_at":"2026-07-05T08:23:34.414766+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2405.16714","created_at":"2026-07-05T08:23:34.414766+00:00"},{"alias_kind":"pith_short_12","alias_value":"RNYLAY3ZDL37","created_at":"2026-07-05T08:23:34.414766+00:00"},{"alias_kind":"pith_short_16","alias_value":"RNYLAY3ZDL375XY2","created_at":"2026-07-05T08:23:34.414766+00:00"},{"alias_kind":"pith_short_8","alias_value":"RNYLAY3Z","created_at":"2026-07-05T08:23:34.414766+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2605.14169","citing_title":"BOOKMARKS: Efficient Active Storyline Memory for Role-playing","ref_index":51,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/RNYLAY3ZDL375XY2DUFK5EXNPQ","json":"https://pith.science/pith/RNYLAY3ZDL375XY2DUFK5EXNPQ.json","graph_json":"https://pith.science/api/pith-number/RNYLAY3ZDL375XY2DUFK5EXNPQ/graph.json","events_json":"https://pith.science/api/pith-number/RNYLAY3ZDL375XY2DUFK5EXNPQ/events.json","paper":"https://pith.science/paper/RNYLAY3Z"},"agent_actions":{"view_html":"https://pith.science/pith/RNYLAY3ZDL375XY2DUFK5EXNPQ","download_json":"https://pith.science/pith/RNYLAY3ZDL375XY2DUFK5EXNPQ.json","view_paper":"https://pith.science/paper/RNYLAY3Z","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2405.16714&json=true","fetch_graph":"https://pith.science/api/pith-number/RNYLAY3ZDL375XY2DUFK5EXNPQ/graph.json","fetch_events":"https://pith.science/api/pith-number/RNYLAY3ZDL375XY2DUFK5EXNPQ/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/RNYLAY3ZDL375XY2DUFK5EXNPQ/action/timestamp_anchor","attest_storage":"https://pith.science/pith/RNYLAY3ZDL375XY2DUFK5EXNPQ/action/storage_attestation","attest_author":"https://pith.science/pith/RNYLAY3ZDL375XY2DUFK5EXNPQ/action/author_attestation","sign_citation":"https://pith.science/pith/RNYLAY3ZDL375XY2DUFK5EXNPQ/action/citation_signature","submit_replication":"https://pith.science/pith/RNYLAY3ZDL375XY2DUFK5EXNPQ/action/replication_record"}},"created_at":"2026-07-05T08:23:34.414766+00:00","updated_at":"2026-07-05T08:23:34.414766+00:00"}