{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:L33RNGUYBWS576LT36D4GTW3WL","short_pith_number":"pith:L33RNGUY","schema_version":"1.0","canonical_sha256":"5ef7169a980da5dff973df87c34edbb2f1730eddd9597ab8b78e8106ed54ac93","source":{"kind":"arxiv","id":"2605.00994","version":2},"attestation_state":"computed","paper":{"title":"Most Current Model Organisms Are Leaky: Perplexity Differencing Often Reveals Finetuning Objectives","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","headline":"Perplexity differencing on completions from random prefills surfaces finetuning objectives in most model organisms.","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Dan Wilhelm, Luca Baroni, Mohammed Abu Baker","submitted_at":"2026-05-01T18:00:55Z","abstract_excerpt":"Finetuning can significantly modify the behavior of large language models, including introducing harmful or unsafe behaviors. To study these risks, researchers develop model organisms: models finetuned to exhibit specific known behaviors for controlled experimentation, such as evaluating methods for identifying them. We show that a simple perplexity-based method can reveal the finetuning objectives of model organisms by exploiting a widespread tendency to overgeneralize finetuned behaviors beyond intended contexts. We generate diverse completions from the finetuned model using short random pre"},"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":"2605.00994","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.CL","submitted_at":"2026-05-01T18:00:55Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"bbeb7847e2c0218b82d28e816a5d6a57d3dc09a8f6399c5e4101c304f0ba6784","abstract_canon_sha256":"98133f29ab394c6fc1eb1e7453dce3e825395a16bc5f801aae67f197ba48a623"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-30T02:18:08.785140Z","signature_b64":"FDnnDkKlHj6PqQ1Px4rMGbyqRz8Obv2WAaArrr51lg6kNxhYctWajb14MeXahWVmK0Zksn8rMnw1HfP628mpDQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"5ef7169a980da5dff973df87c34edbb2f1730eddd9597ab8b78e8106ed54ac93","last_reissued_at":"2026-06-30T02:18:08.784645Z","signature_status":"signed_v1","first_computed_at":"2026-06-30T02:18:08.784645Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Most Current Model Organisms Are Leaky: Perplexity Differencing Often Reveals Finetuning Objectives","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","headline":"Perplexity differencing on completions from random prefills surfaces finetuning objectives in most model organisms.","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Dan Wilhelm, Luca Baroni, Mohammed Abu Baker","submitted_at":"2026-05-01T18:00:55Z","abstract_excerpt":"Finetuning can significantly modify the behavior of large language models, including introducing harmful or unsafe behaviors. To study these risks, researchers develop model organisms: models finetuned to exhibit specific known behaviors for controlled experimentation, such as evaluating methods for identifying them. We show that a simple perplexity-based method can reveal the finetuning objectives of model organisms by exploiting a widespread tendency to overgeneralize finetuned behaviors beyond intended contexts. We generate diverse completions from the finetuned model using short random pre"},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"a simple perplexity-based method can surface finetuning objectives from model organisms by leveraging their tendency to overgeneralize their finetuned behaviors beyond the intended context","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"that finetuned models reliably overgeneralize their training objectives to unrelated contexts generated from random prefills, producing detectable perplexity gaps against a reference model","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"Perplexity gaps between finetuned and reference models on random-prefill completions often reveal the original finetuning objectives across diverse model organisms.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"Perplexity differencing on completions from random prefills surfaces finetuning objectives in most model organisms.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"92ed26ec09b935a7177b025b6539c412aabe204323ca7a3a900a9dbf28128d73"},"source":{"id":"2605.00994","kind":"arxiv","version":2},"verdict":{"id":"1f349c81-ec89-41f7-be14-e19108758f33","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-09T18:43:49.422700Z","strongest_claim":"a simple perplexity-based method can surface finetuning objectives from model organisms by leveraging their tendency to overgeneralize their finetuned behaviors beyond the intended context","one_line_summary":"Perplexity gaps between finetuned and reference models on random-prefill completions often reveal the original finetuning objectives across diverse model organisms.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"that finetuned models reliably overgeneralize their training objectives to unrelated contexts generated from random prefills, producing detectable perplexity gaps against a reference model","pith_extraction_headline":"Perplexity differencing on completions from random prefills surfaces finetuning objectives in most model organisms."},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2605.00994/integrity.json","findings":[],"available":true,"detectors_run":[{"name":"ai_meta_artifact","ran_at":"2026-05-20T18:40:27.363128Z","status":"completed","version":"1.0.0","findings_count":0},{"name":"doi_compliance","ran_at":"2026-05-19T17:45:49.779381Z","status":"completed","version":"1.0.0","findings_count":0}],"snapshot_sha256":"5e4c80c66af0ca401df3410dd2b1dc3fd6c8d168c508588a3c422e716c45b319"},"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":"2605.00994","created_at":"2026-06-30T02:18:08.784705+00:00"},{"alias_kind":"arxiv_version","alias_value":"2605.00994v2","created_at":"2026-06-30T02:18:08.784705+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.00994","created_at":"2026-06-30T02:18:08.784705+00:00"},{"alias_kind":"pith_short_12","alias_value":"L33RNGUYBWS5","created_at":"2026-06-30T02:18:08.784705+00:00"},{"alias_kind":"pith_short_16","alias_value":"L33RNGUYBWS576LT","created_at":"2026-06-30T02:18:08.784705+00:00"},{"alias_kind":"pith_short_8","alias_value":"L33RNGUY","created_at":"2026-06-30T02:18:08.784705+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/L33RNGUYBWS576LT36D4GTW3WL","json":"https://pith.science/pith/L33RNGUYBWS576LT36D4GTW3WL.json","graph_json":"https://pith.science/api/pith-number/L33RNGUYBWS576LT36D4GTW3WL/graph.json","events_json":"https://pith.science/api/pith-number/L33RNGUYBWS576LT36D4GTW3WL/events.json","paper":"https://pith.science/paper/L33RNGUY"},"agent_actions":{"view_html":"https://pith.science/pith/L33RNGUYBWS576LT36D4GTW3WL","download_json":"https://pith.science/pith/L33RNGUYBWS576LT36D4GTW3WL.json","view_paper":"https://pith.science/paper/L33RNGUY","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2605.00994&json=true","fetch_graph":"https://pith.science/api/pith-number/L33RNGUYBWS576LT36D4GTW3WL/graph.json","fetch_events":"https://pith.science/api/pith-number/L33RNGUYBWS576LT36D4GTW3WL/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/L33RNGUYBWS576LT36D4GTW3WL/action/timestamp_anchor","attest_storage":"https://pith.science/pith/L33RNGUYBWS576LT36D4GTW3WL/action/storage_attestation","attest_author":"https://pith.science/pith/L33RNGUYBWS576LT36D4GTW3WL/action/author_attestation","sign_citation":"https://pith.science/pith/L33RNGUYBWS576LT36D4GTW3WL/action/citation_signature","submit_replication":"https://pith.science/pith/L33RNGUYBWS576LT36D4GTW3WL/action/replication_record"}},"created_at":"2026-06-30T02:18:08.784705+00:00","updated_at":"2026-06-30T02:18:08.784705+00:00"}