{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:YANKGR7A4XT64ZDSR2RBVV6A63","short_pith_number":"pith:YANKGR7A","schema_version":"1.0","canonical_sha256":"c01aa347e0e5e7ee64728ea21ad7c0f6e394d77e9f504f0bd1c8a90b63c49c2c","source":{"kind":"arxiv","id":"2605.23925","version":1},"attestation_state":"computed","paper":{"title":"Catching The Correct Answer Trap: Characterising AI Tutor Blind Spots When Analysing Student Reasoning","license":"http://creativecommons.org/licenses/by-sa/4.0/","headline":"","cross_cats":["cs.AI","cs.CL"],"primary_cat":"cs.CY","authors_text":"Moiz Imran, Sahan Bulathwela","submitted_at":"2026-04-20T10:53:48Z","abstract_excerpt":"Intelligent tutoring systems increasingly provide automated feedback on student work, but robust feedback requires assessing reasoning, not only final answers. We study a failure mode we call the correct answer trap (CAT): models under-detect misconceptions when students reach a correct answer via flawed reasoning. Analysing real student responses from the Eedi mathematics platform, we show that 71% of these failures concentrate in just two question types, both sharing a common structure where flawed reasoning happens to produce the correct numerical answer. Comparing a fine-tuned T5 with a fr"},"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":"2605.23925","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by-sa/4.0/","primary_cat":"cs.CY","submitted_at":"2026-04-20T10:53:48Z","cross_cats_sorted":["cs.AI","cs.CL"],"title_canon_sha256":"5c3c814797d299d2699cbe46b26bb4f4a90015a0bdcf5f59cc634979c002d31d","abstract_canon_sha256":"6df2acdd4969dc001f32aa3378e0f6d1b982b9387341d5a743220714644acf2e"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-26T00:02:22.218398Z","signature_b64":"RewP0aiDywWO0FVJShJ8YSXAtSWmDqx+IJ3GkHOwSA2LjTKU0nOAWpeAqCupOPhX2omL8EO8oDCtZzcW1q4hDw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"c01aa347e0e5e7ee64728ea21ad7c0f6e394d77e9f504f0bd1c8a90b63c49c2c","last_reissued_at":"2026-05-26T00:02:22.217287Z","signature_status":"signed_v1","first_computed_at":"2026-05-26T00:02:22.217287Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Catching The Correct Answer Trap: Characterising AI Tutor Blind Spots When Analysing Student Reasoning","license":"http://creativecommons.org/licenses/by-sa/4.0/","headline":"","cross_cats":["cs.AI","cs.CL"],"primary_cat":"cs.CY","authors_text":"Moiz Imran, Sahan Bulathwela","submitted_at":"2026-04-20T10:53:48Z","abstract_excerpt":"Intelligent tutoring systems increasingly provide automated feedback on student work, but robust feedback requires assessing reasoning, not only final answers. We study a failure mode we call the correct answer trap (CAT): models under-detect misconceptions when students reach a correct answer via flawed reasoning. Analysing real student responses from the Eedi mathematics platform, we show that 71% of these failures concentrate in just two question types, both sharing a common structure where flawed reasoning happens to produce the correct numerical answer. Comparing a fine-tuned T5 with a fr"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2605.23925","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/2605.23925/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":"2605.23925","created_at":"2026-05-26T00:02:22.217418+00:00"},{"alias_kind":"arxiv_version","alias_value":"2605.23925v1","created_at":"2026-05-26T00:02:22.217418+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.23925","created_at":"2026-05-26T00:02:22.217418+00:00"},{"alias_kind":"pith_short_12","alias_value":"YANKGR7A4XT6","created_at":"2026-05-26T00:02:22.217418+00:00"},{"alias_kind":"pith_short_16","alias_value":"YANKGR7A4XT64ZDS","created_at":"2026-05-26T00:02:22.217418+00:00"},{"alias_kind":"pith_short_8","alias_value":"YANKGR7A","created_at":"2026-05-26T00:02:22.217418+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":0,"internal_anchor_count":0,"sample":[]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/YANKGR7A4XT64ZDSR2RBVV6A63","json":"https://pith.science/pith/YANKGR7A4XT64ZDSR2RBVV6A63.json","graph_json":"https://pith.science/api/pith-number/YANKGR7A4XT64ZDSR2RBVV6A63/graph.json","events_json":"https://pith.science/api/pith-number/YANKGR7A4XT64ZDSR2RBVV6A63/events.json","paper":"https://pith.science/paper/YANKGR7A"},"agent_actions":{"view_html":"https://pith.science/pith/YANKGR7A4XT64ZDSR2RBVV6A63","download_json":"https://pith.science/pith/YANKGR7A4XT64ZDSR2RBVV6A63.json","view_paper":"https://pith.science/paper/YANKGR7A","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2605.23925&json=true","fetch_graph":"https://pith.science/api/pith-number/YANKGR7A4XT64ZDSR2RBVV6A63/graph.json","fetch_events":"https://pith.science/api/pith-number/YANKGR7A4XT64ZDSR2RBVV6A63/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/YANKGR7A4XT64ZDSR2RBVV6A63/action/timestamp_anchor","attest_storage":"https://pith.science/pith/YANKGR7A4XT64ZDSR2RBVV6A63/action/storage_attestation","attest_author":"https://pith.science/pith/YANKGR7A4XT64ZDSR2RBVV6A63/action/author_attestation","sign_citation":"https://pith.science/pith/YANKGR7A4XT64ZDSR2RBVV6A63/action/citation_signature","submit_replication":"https://pith.science/pith/YANKGR7A4XT64ZDSR2RBVV6A63/action/replication_record"}},"created_at":"2026-05-26T00:02:22.217418+00:00","updated_at":"2026-05-26T00:02:22.217418+00:00"}