{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2024:DD52LTREYE4RDLAKFUDYM6453Z","short_pith_number":"pith:DD52LTRE","canonical_record":{"source":{"id":"2407.12847","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-07-05T09:26:40Z","cross_cats_sorted":["cs.AI","cs.HC"],"title_canon_sha256":"375c04ed5b44422a4a0a5b2ef183526687e50f9a9e76bba5146f1e039da0a927","abstract_canon_sha256":"bbff29aa5acf06fd22019f528fb902b9ae99bf7a2ef431c7859da68604110c46"},"schema_version":"1.0"},"canonical_sha256":"18fba5ce24c13911ac0a2d07867b9dde7335a00ef1e94004e31797d8b52c640f","source":{"kind":"arxiv","id":"2407.12847","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2407.12847","created_at":"2026-07-05T08:45:12Z"},{"alias_kind":"arxiv_version","alias_value":"2407.12847v1","created_at":"2026-07-05T08:45:12Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2407.12847","created_at":"2026-07-05T08:45:12Z"},{"alias_kind":"pith_short_12","alias_value":"DD52LTREYE4R","created_at":"2026-07-05T08:45:12Z"},{"alias_kind":"pith_short_16","alias_value":"DD52LTREYE4RDLAK","created_at":"2026-07-05T08:45:12Z"},{"alias_kind":"pith_short_8","alias_value":"DD52LTRE","created_at":"2026-07-05T08:45:12Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2024:DD52LTREYE4RDLAKFUDYM6453Z","target":"record","payload":{"canonical_record":{"source":{"id":"2407.12847","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-07-05T09:26:40Z","cross_cats_sorted":["cs.AI","cs.HC"],"title_canon_sha256":"375c04ed5b44422a4a0a5b2ef183526687e50f9a9e76bba5146f1e039da0a927","abstract_canon_sha256":"bbff29aa5acf06fd22019f528fb902b9ae99bf7a2ef431c7859da68604110c46"},"schema_version":"1.0"},"canonical_sha256":"18fba5ce24c13911ac0a2d07867b9dde7335a00ef1e94004e31797d8b52c640f","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T08:45:12.478417Z","signature_b64":"5yUfWE4USczhA8ZCJUp00G02d32L1hBBSLvkxTI2Xu9Rc9CD5J/PjrQJEt56XfLz9GcSk7QHGnD3q5rpOHNuDg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"18fba5ce24c13911ac0a2d07867b9dde7335a00ef1e94004e31797d8b52c640f","last_reissued_at":"2026-07-05T08:45:12.478013Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T08:45:12.478013Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2407.12847","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":"2026-07-05T08:45:12Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"MyxQxS1akMOgBJ0hRbK1DpaU3Dut5+1JdL5lGQ9n0ProunqCXDHd8aMPZabyYnh3UM2aBBnil/1gf/WADTkXAA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-06T08:35:00.155863Z"},"content_sha256":"d21a4a966142eb4c067d809f82f4a78a98d256ba234739d57ae8b17e9736d909","schema_version":"1.0","event_id":"sha256:d21a4a966142eb4c067d809f82f4a78a98d256ba234739d57ae8b17e9736d909"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2024:DD52LTREYE4RDLAKFUDYM6453Z","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Aligning Model Evaluations with Human Preferences: Mitigating Token Count Bias in Language Model Assessments","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.HC"],"primary_cat":"cs.CL","authors_text":"Jason Mars, Roland Daynauth","submitted_at":"2024-07-05T09:26:40Z","abstract_excerpt":"The SLAM paper demonstrated that on-device Small Language Models (SLMs) are a viable and cost-effective alternative to API-based Large Language Models (LLMs), such as OpenAI's GPT-4, offering comparable performance and stability. However, SLAM also identified discrepancies between human preferences and traditional auto-evaluators. This follow-up paper explores methods to align LLM evaluator preferences with human evaluations by addressing biases, particularly toward higher token counts. We employed Bayesian statistics and a t-test to quantify this bias and developed a recalibration procedure t"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2407.12847","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/2407.12847/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":"2026-07-05T08:45:12Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"PkTDb5oHIaJAcSjIIcXPTx2jGQfJhMeEZfFlNgGRtKwyXMe4bN8Z1yRCEzHfLQkszrW4mtl9si7UlepxHceaBg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-06T08:35:00.157031Z"},"content_sha256":"28bda40ac8c677f6e4162fc10b5c71426087cd5528cdb2fd03231d5b05ba081f","schema_version":"1.0","event_id":"sha256:28bda40ac8c677f6e4162fc10b5c71426087cd5528cdb2fd03231d5b05ba081f"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/DD52LTREYE4RDLAKFUDYM6453Z/bundle.json","state_url":"https://pith.science/pith/DD52LTREYE4RDLAKFUDYM6453Z/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/DD52LTREYE4RDLAKFUDYM6453Z/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":"2026-08-06T08:35:00Z","links":{"resolver":"https://pith.science/pith/DD52LTREYE4RDLAKFUDYM6453Z","bundle":"https://pith.science/pith/DD52LTREYE4RDLAKFUDYM6453Z/bundle.json","state":"https://pith.science/pith/DD52LTREYE4RDLAKFUDYM6453Z/state.json","well_known_bundle":"https://pith.science/.well-known/pith/DD52LTREYE4RDLAKFUDYM6453Z/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2024:DD52LTREYE4RDLAKFUDYM6453Z","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":"bbff29aa5acf06fd22019f528fb902b9ae99bf7a2ef431c7859da68604110c46","cross_cats_sorted":["cs.AI","cs.HC"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-07-05T09:26:40Z","title_canon_sha256":"375c04ed5b44422a4a0a5b2ef183526687e50f9a9e76bba5146f1e039da0a927"},"schema_version":"1.0","source":{"id":"2407.12847","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2407.12847","created_at":"2026-07-05T08:45:12Z"},{"alias_kind":"arxiv_version","alias_value":"2407.12847v1","created_at":"2026-07-05T08:45:12Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2407.12847","created_at":"2026-07-05T08:45:12Z"},{"alias_kind":"pith_short_12","alias_value":"DD52LTREYE4R","created_at":"2026-07-05T08:45:12Z"},{"alias_kind":"pith_short_16","alias_value":"DD52LTREYE4RDLAK","created_at":"2026-07-05T08:45:12Z"},{"alias_kind":"pith_short_8","alias_value":"DD52LTRE","created_at":"2026-07-05T08:45:12Z"}],"graph_snapshots":[{"event_id":"sha256:28bda40ac8c677f6e4162fc10b5c71426087cd5528cdb2fd03231d5b05ba081f","target":"graph","created_at":"2026-07-05T08:45:12Z","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/2407.12847/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"The SLAM paper demonstrated that on-device Small Language Models (SLMs) are a viable and cost-effective alternative to API-based Large Language Models (LLMs), such as OpenAI's GPT-4, offering comparable performance and stability. However, SLAM also identified discrepancies between human preferences and traditional auto-evaluators. This follow-up paper explores methods to align LLM evaluator preferences with human evaluations by addressing biases, particularly toward higher token counts. We employed Bayesian statistics and a t-test to quantify this bias and developed a recalibration procedure t","authors_text":"Jason Mars, Roland Daynauth","cross_cats":["cs.AI","cs.HC"],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-07-05T09:26:40Z","title":"Aligning Model Evaluations with Human Preferences: Mitigating Token Count Bias in Language Model Assessments"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2407.12847","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:d21a4a966142eb4c067d809f82f4a78a98d256ba234739d57ae8b17e9736d909","target":"record","created_at":"2026-07-05T08:45:12Z","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":"bbff29aa5acf06fd22019f528fb902b9ae99bf7a2ef431c7859da68604110c46","cross_cats_sorted":["cs.AI","cs.HC"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-07-05T09:26:40Z","title_canon_sha256":"375c04ed5b44422a4a0a5b2ef183526687e50f9a9e76bba5146f1e039da0a927"},"schema_version":"1.0","source":{"id":"2407.12847","kind":"arxiv","version":1}},"canonical_sha256":"18fba5ce24c13911ac0a2d07867b9dde7335a00ef1e94004e31797d8b52c640f","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"18fba5ce24c13911ac0a2d07867b9dde7335a00ef1e94004e31797d8b52c640f","first_computed_at":"2026-07-05T08:45:12.478013Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T08:45:12.478013Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"5yUfWE4USczhA8ZCJUp00G02d32L1hBBSLvkxTI2Xu9Rc9CD5J/PjrQJEt56XfLz9GcSk7QHGnD3q5rpOHNuDg==","signature_status":"signed_v1","signed_at":"2026-07-05T08:45:12.478417Z","signed_message":"canonical_sha256_bytes"},"source_id":"2407.12847","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:d21a4a966142eb4c067d809f82f4a78a98d256ba234739d57ae8b17e9736d909","sha256:28bda40ac8c677f6e4162fc10b5c71426087cd5528cdb2fd03231d5b05ba081f"],"state_sha256":"718b00b37afb83e042776b2cc34d6dd47e2b162f29d6c14060898666c09a5954"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"uA6VkVuDQkJqUIO6I0+WeQOHmVvdiezI9/nS85VEKasyig/wjGSD4I279Q1wgau7zZJ6gTkSdNGwwW3oZYKhCw==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-06T08:35:00.167479Z","bundle_sha256":"c55d031d8f79cd9082e898bf16de3a0271385838f550cc776e14bbd0d9d25c30"}}