{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2025:ZG4YTXHAYRJVZJGLJKYSKY457T","short_pith_number":"pith:ZG4YTXHA","canonical_record":{"source":{"id":"2507.05266","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-06-30T06:14:32Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"c7578d00c1200f5c9edca9a710849b2dcfdd7798a189239ed5160c113eb31bb6","abstract_canon_sha256":"6731a3ee4f6e0f792661037c3869445f811daa850b0f9d87d2f5aabfefda8331"},"schema_version":"1.0"},"canonical_sha256":"c9b989dce0c4535ca4cb4ab125639dfcf9d94522f382f2f2b0e78cc521c8274c","source":{"kind":"arxiv","id":"2507.05266","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2507.05266","created_at":"2026-07-05T11:33:27Z"},{"alias_kind":"arxiv_version","alias_value":"2507.05266v1","created_at":"2026-07-05T11:33:27Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2507.05266","created_at":"2026-07-05T11:33:27Z"},{"alias_kind":"pith_short_12","alias_value":"ZG4YTXHAYRJV","created_at":"2026-07-05T11:33:27Z"},{"alias_kind":"pith_short_16","alias_value":"ZG4YTXHAYRJVZJGL","created_at":"2026-07-05T11:33:27Z"},{"alias_kind":"pith_short_8","alias_value":"ZG4YTXHA","created_at":"2026-07-05T11:33:27Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2025:ZG4YTXHAYRJVZJGLJKYSKY457T","target":"record","payload":{"canonical_record":{"source":{"id":"2507.05266","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-06-30T06:14:32Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"c7578d00c1200f5c9edca9a710849b2dcfdd7798a189239ed5160c113eb31bb6","abstract_canon_sha256":"6731a3ee4f6e0f792661037c3869445f811daa850b0f9d87d2f5aabfefda8331"},"schema_version":"1.0"},"canonical_sha256":"c9b989dce0c4535ca4cb4ab125639dfcf9d94522f382f2f2b0e78cc521c8274c","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:33:27.302590Z","signature_b64":"4jOai5d4qBjF1DW+5juQWDWykcSeCJ5f9lnLtCFXWU32HKn5zuSX3OFyZ1vs2nh8MTrLfd8j4edJXep9MsX+BQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"c9b989dce0c4535ca4cb4ab125639dfcf9d94522f382f2f2b0e78cc521c8274c","last_reissued_at":"2026-07-05T11:33:27.302122Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:33:27.302122Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2507.05266","source_version":1,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T11:33:27Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"p986bLEyXowLE+lggQpoTJQ8TKQGXGqO81ooUwY83BTiipKbd2IDhLdGAn270rn7Zo9ez2HuW5vZLPRzsv4VCw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-10T10:17:53.707677Z"},"content_sha256":"98e64c07097cac88ba4f263533b542e536c20a7198c54b78fc7e7280481ef68f","schema_version":"1.0","event_id":"sha256:98e64c07097cac88ba4f263533b542e536c20a7198c54b78fc7e7280481ef68f"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2025:ZG4YTXHAYRJVZJGLJKYSKY457T","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"User Behavior Prediction as a Generic, Robust, Scalable, and Low-Cost Evaluation Strategy for Estimating Generalization in LLMs","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Monojit Choudhury, Sougata Saha","submitted_at":"2025-06-30T06:14:32Z","abstract_excerpt":"Measuring the generalization ability of Large Language Models (LLMs) is challenging due to data contamination. As models grow and computation becomes cheaper, ensuring tasks and test cases are unseen during training phases will become nearly impossible. We argue that knowledge-retrieval and reasoning tasks are not ideal for measuring generalization, as LLMs are not trained for specific tasks. Instead, we propose user behavior prediction, also a key aspect of personalization, as a theoretically sound, scalable, and robust alternative. We introduce a novel framework for this approach and test it"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2507.05266","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/2507.05266/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"},"verdict_id":null},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T11:33:27Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"tR7SzeBU/i67B6GUeQ+1b1PVa+Y62HQuFyLwTRfEXiOk4U7DbBSdMnv6lhawR/WunfJkq97JXQxGyYWifhVDAA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-10T10:17:53.708185Z"},"content_sha256":"9dff5777892a9fea5e71809af8d4c8a6abeffc264794eeac81efbfd7ae4f9dc1","schema_version":"1.0","event_id":"sha256:9dff5777892a9fea5e71809af8d4c8a6abeffc264794eeac81efbfd7ae4f9dc1"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/ZG4YTXHAYRJVZJGLJKYSKY457T/bundle.json","state_url":"https://pith.science/pith/ZG4YTXHAYRJVZJGLJKYSKY457T/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/ZG4YTXHAYRJVZJGLJKYSKY457T/bundle.json","status":"primary"}],"public_keys":[{"key_id":"pith-v1-2026-05","algorithm":"ed25519","format":"raw","public_key_b64":"stVStoiQhXFxp4s2pdzPNoqVNBMojDU/fJ2db5S3CbM=","public_key_hex":"b2d552b68890857171a78b36a5dccf368a953413288c353f7c9d9d6f94b709b3","fingerprint_sha256_b32_first128bits":"RVFV5Z2OI2J3ZUO7ERDEBCYNKS","fingerprint_sha256_hex":"8d4b5ee74e4693bcd1df2446408b0d54","rotates_at":null,"url":"https://pith.science/pith-signing-key.json","notes":"Pith uses this Ed25519 key to sign canonical record SHA-256 digests. Verify with: ed25519_verify(public_key, message=canonical_sha256_bytes, signature=base64decode(signature_b64))."}],"merge_version":"pith-open-graph-merge-v1","built_at":"2026-08-10T10:17:53Z","links":{"resolver":"https://pith.science/pith/ZG4YTXHAYRJVZJGLJKYSKY457T","bundle":"https://pith.science/pith/ZG4YTXHAYRJVZJGLJKYSKY457T/bundle.json","state":"https://pith.science/pith/ZG4YTXHAYRJVZJGLJKYSKY457T/state.json","well_known_bundle":"https://pith.science/.well-known/pith/ZG4YTXHAYRJVZJGLJKYSKY457T/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2025:ZG4YTXHAYRJVZJGLJKYSKY457T","merge_version":"pith-open-graph-merge-v1","event_count":2,"valid_event_count":2,"invalid_event_count":0,"equivocation_count":0,"current":{"canonical_record":{"metadata":{"abstract_canon_sha256":"6731a3ee4f6e0f792661037c3869445f811daa850b0f9d87d2f5aabfefda8331","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-06-30T06:14:32Z","title_canon_sha256":"c7578d00c1200f5c9edca9a710849b2dcfdd7798a189239ed5160c113eb31bb6"},"schema_version":"1.0","source":{"id":"2507.05266","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2507.05266","created_at":"2026-07-05T11:33:27Z"},{"alias_kind":"arxiv_version","alias_value":"2507.05266v1","created_at":"2026-07-05T11:33:27Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2507.05266","created_at":"2026-07-05T11:33:27Z"},{"alias_kind":"pith_short_12","alias_value":"ZG4YTXHAYRJV","created_at":"2026-07-05T11:33:27Z"},{"alias_kind":"pith_short_16","alias_value":"ZG4YTXHAYRJVZJGL","created_at":"2026-07-05T11:33:27Z"},{"alias_kind":"pith_short_8","alias_value":"ZG4YTXHA","created_at":"2026-07-05T11:33:27Z"}],"graph_snapshots":[{"event_id":"sha256:9dff5777892a9fea5e71809af8d4c8a6abeffc264794eeac81efbfd7ae4f9dc1","target":"graph","created_at":"2026-07-05T11:33:27Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"graph_snapshot":{"author_claims":{"count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","strong_count":0},"builder_version":"pith-number-builder-2026-05-17-v1","claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"integrity":{"available":true,"clean":true,"detectors_run":[],"endpoint":"/pith/2507.05266/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Measuring the generalization ability of Large Language Models (LLMs) is challenging due to data contamination. As models grow and computation becomes cheaper, ensuring tasks and test cases are unseen during training phases will become nearly impossible. We argue that knowledge-retrieval and reasoning tasks are not ideal for measuring generalization, as LLMs are not trained for specific tasks. Instead, we propose user behavior prediction, also a key aspect of personalization, as a theoretically sound, scalable, and robust alternative. We introduce a novel framework for this approach and test it","authors_text":"Monojit Choudhury, Sougata Saha","cross_cats":["cs.AI"],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-06-30T06:14:32Z","title":"User Behavior Prediction as a Generic, Robust, Scalable, and Low-Cost Evaluation Strategy for Estimating Generalization in LLMs"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2507.05266","kind":"arxiv","version":1},"verdict":{"created_at":null,"id":null,"model_set":{},"one_line_summary":"","pipeline_version":null,"pith_extraction_headline":"","strongest_claim":"","weakest_assumption":""}},"verdict_id":null}}],"author_attestations":[],"timestamp_anchors":[],"storage_attestations":[],"citation_signatures":[],"replication_records":[],"corrections":[],"mirror_hints":[],"record_created":{"event_id":"sha256:98e64c07097cac88ba4f263533b542e536c20a7198c54b78fc7e7280481ef68f","target":"record","created_at":"2026-07-05T11:33:27Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"attestation_state":"computed","canonical_record":{"metadata":{"abstract_canon_sha256":"6731a3ee4f6e0f792661037c3869445f811daa850b0f9d87d2f5aabfefda8331","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-06-30T06:14:32Z","title_canon_sha256":"c7578d00c1200f5c9edca9a710849b2dcfdd7798a189239ed5160c113eb31bb6"},"schema_version":"1.0","source":{"id":"2507.05266","kind":"arxiv","version":1}},"canonical_sha256":"c9b989dce0c4535ca4cb4ab125639dfcf9d94522f382f2f2b0e78cc521c8274c","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"c9b989dce0c4535ca4cb4ab125639dfcf9d94522f382f2f2b0e78cc521c8274c","first_computed_at":"2026-07-05T11:33:27.302122Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T11:33:27.302122Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"4jOai5d4qBjF1DW+5juQWDWykcSeCJ5f9lnLtCFXWU32HKn5zuSX3OFyZ1vs2nh8MTrLfd8j4edJXep9MsX+BQ==","signature_status":"signed_v1","signed_at":"2026-07-05T11:33:27.302590Z","signed_message":"canonical_sha256_bytes"},"source_id":"2507.05266","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:98e64c07097cac88ba4f263533b542e536c20a7198c54b78fc7e7280481ef68f","sha256:9dff5777892a9fea5e71809af8d4c8a6abeffc264794eeac81efbfd7ae4f9dc1"],"state_sha256":"4bf7dcad5a1372336844502365cd93abcb2fc80da160fd7aa9be6b3f1a958e71"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"nEH2GtuZ3eQAurUwqHXwAYAa9Os1s9Q5zXOwmV3osl7x70Av9V6a+ESQPa8ERF01GHDH48r57EZjVJubUdyJAg==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-10T10:17:53.713934Z","bundle_sha256":"2674ad4ba1aeb9f8b181b3a1a334b5bcb8996b7cf7bca3739a7bc321e972f0c0"}}