{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2025:H34GUXUUCLZJ2LYJJPVSXOI7WH","short_pith_number":"pith:H34GUXUU","canonical_record":{"source":{"id":"2501.12147","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2025-01-21T14:00:43Z","cross_cats_sorted":["cs.AI","cs.LG"],"title_canon_sha256":"13de072c8a8453658eaa2489c9b4f3c113251f53f2c11977e6ceec94fda31806","abstract_canon_sha256":"6990a95925d07e392e234d702db5148f46979be4a1c4c905af6b74d54161da0c"},"schema_version":"1.0"},"canonical_sha256":"3ef86a5e9412f29d2f094beb2bb91fb1e9cb1bfb9566edce49a15382b527f947","source":{"kind":"arxiv","id":"2501.12147","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2501.12147","created_at":"2026-07-05T10:03:32Z"},{"alias_kind":"arxiv_version","alias_value":"2501.12147v1","created_at":"2026-07-05T10:03:32Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2501.12147","created_at":"2026-07-05T10:03:32Z"},{"alias_kind":"pith_short_12","alias_value":"H34GUXUUCLZJ","created_at":"2026-07-05T10:03:32Z"},{"alias_kind":"pith_short_16","alias_value":"H34GUXUUCLZJ2LYJ","created_at":"2026-07-05T10:03:32Z"},{"alias_kind":"pith_short_8","alias_value":"H34GUXUU","created_at":"2026-07-05T10:03:32Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2025:H34GUXUUCLZJ2LYJJPVSXOI7WH","target":"record","payload":{"canonical_record":{"source":{"id":"2501.12147","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2025-01-21T14:00:43Z","cross_cats_sorted":["cs.AI","cs.LG"],"title_canon_sha256":"13de072c8a8453658eaa2489c9b4f3c113251f53f2c11977e6ceec94fda31806","abstract_canon_sha256":"6990a95925d07e392e234d702db5148f46979be4a1c4c905af6b74d54161da0c"},"schema_version":"1.0"},"canonical_sha256":"3ef86a5e9412f29d2f094beb2bb91fb1e9cb1bfb9566edce49a15382b527f947","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:03:32.025188Z","signature_b64":"v9Dz1q9E28z4CDdMpF5S9yZRjm7T3qvRuWYp/yepmO3eUt0R6NGZBFjgzgJCy0D0dHSiOxn6rE25d9yeG+gSBQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"3ef86a5e9412f29d2f094beb2bb91fb1e9cb1bfb9566edce49a15382b527f947","last_reissued_at":"2026-07-05T10:03:32.024780Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:03:32.024780Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2501.12147","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":"2026-07-05T10:03:32Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"6q03rK8Ee7VK23c4/pXwn7DX9EaJAT4FfpVdYcBswC9Hhs/wRIATaIXh6VuBw50vJERnEDWG/qRof6xd+L0YBQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-10T03:14:51.025194Z"},"content_sha256":"7a5f677aac2d1c7aa3c9eeedbe3ff880a46be180d34f23c849b3a31733025fb0","schema_version":"1.0","event_id":"sha256:7a5f677aac2d1c7aa3c9eeedbe3ff880a46be180d34f23c849b3a31733025fb0"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2025:H34GUXUUCLZJ2LYJJPVSXOI7WH","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Improving Influence-based Instruction Tuning Data Selection for Balanced Learning of Diverse Capabilities","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.LG"],"primary_cat":"cs.CL","authors_text":"Dylan Zhang, Hao Peng, Jiaqi W. Ma, Qirun Dai","submitted_at":"2025-01-21T14:00:43Z","abstract_excerpt":"Selecting appropriate training data is crucial for effective instruction fine-tuning of large language models (LLMs), which aims to (1) elicit strong capabilities, and (2) achieve balanced performance across a diverse range of tasks. Influence-based methods show promise in achieving (1) by estimating the contribution of each training example to the model's predictions, but often struggle with (2). Our systematic investigation reveals that this underperformance can be attributed to an inherent bias where certain tasks intrinsically have greater influence than others. As a result, data selection"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2501.12147","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/2501.12147/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":"2026-07-05T10:03:32Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"XzYO06JlnDCJ23VloxE1flIHhdM1lbDXCmrKtsJF4pSzQujAm/EpnBsMk8nT/Oj+jovnYJza13ZWu3Lpy631AQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-10T03:14:51.025704Z"},"content_sha256":"26e676c8cb847f25eb82594631fb65028992dfc461a45a789a14831e13188912","schema_version":"1.0","event_id":"sha256:26e676c8cb847f25eb82594631fb65028992dfc461a45a789a14831e13188912"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/H34GUXUUCLZJ2LYJJPVSXOI7WH/bundle.json","state_url":"https://pith.science/pith/H34GUXUUCLZJ2LYJJPVSXOI7WH/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/H34GUXUUCLZJ2LYJJPVSXOI7WH/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":"2026-08-10T03:14:51Z","links":{"resolver":"https://pith.science/pith/H34GUXUUCLZJ2LYJJPVSXOI7WH","bundle":"https://pith.science/pith/H34GUXUUCLZJ2LYJJPVSXOI7WH/bundle.json","state":"https://pith.science/pith/H34GUXUUCLZJ2LYJJPVSXOI7WH/state.json","well_known_bundle":"https://pith.science/.well-known/pith/H34GUXUUCLZJ2LYJJPVSXOI7WH/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2025:H34GUXUUCLZJ2LYJJPVSXOI7WH","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":"6990a95925d07e392e234d702db5148f46979be4a1c4c905af6b74d54161da0c","cross_cats_sorted":["cs.AI","cs.LG"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2025-01-21T14:00:43Z","title_canon_sha256":"13de072c8a8453658eaa2489c9b4f3c113251f53f2c11977e6ceec94fda31806"},"schema_version":"1.0","source":{"id":"2501.12147","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2501.12147","created_at":"2026-07-05T10:03:32Z"},{"alias_kind":"arxiv_version","alias_value":"2501.12147v1","created_at":"2026-07-05T10:03:32Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2501.12147","created_at":"2026-07-05T10:03:32Z"},{"alias_kind":"pith_short_12","alias_value":"H34GUXUUCLZJ","created_at":"2026-07-05T10:03:32Z"},{"alias_kind":"pith_short_16","alias_value":"H34GUXUUCLZJ2LYJ","created_at":"2026-07-05T10:03:32Z"},{"alias_kind":"pith_short_8","alias_value":"H34GUXUU","created_at":"2026-07-05T10:03:32Z"}],"graph_snapshots":[{"event_id":"sha256:26e676c8cb847f25eb82594631fb65028992dfc461a45a789a14831e13188912","target":"graph","created_at":"2026-07-05T10:03:32Z","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/2501.12147/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Selecting appropriate training data is crucial for effective instruction fine-tuning of large language models (LLMs), which aims to (1) elicit strong capabilities, and (2) achieve balanced performance across a diverse range of tasks. Influence-based methods show promise in achieving (1) by estimating the contribution of each training example to the model's predictions, but often struggle with (2). Our systematic investigation reveals that this underperformance can be attributed to an inherent bias where certain tasks intrinsically have greater influence than others. As a result, data selection","authors_text":"Dylan Zhang, Hao Peng, Jiaqi W. Ma, Qirun Dai","cross_cats":["cs.AI","cs.LG"],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2025-01-21T14:00:43Z","title":"Improving Influence-based Instruction Tuning Data Selection for Balanced Learning of Diverse Capabilities"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2501.12147","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":"sha256:7a5f677aac2d1c7aa3c9eeedbe3ff880a46be180d34f23c849b3a31733025fb0","target":"record","created_at":"2026-07-05T10:03:32Z","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":"6990a95925d07e392e234d702db5148f46979be4a1c4c905af6b74d54161da0c","cross_cats_sorted":["cs.AI","cs.LG"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2025-01-21T14:00:43Z","title_canon_sha256":"13de072c8a8453658eaa2489c9b4f3c113251f53f2c11977e6ceec94fda31806"},"schema_version":"1.0","source":{"id":"2501.12147","kind":"arxiv","version":1}},"canonical_sha256":"3ef86a5e9412f29d2f094beb2bb91fb1e9cb1bfb9566edce49a15382b527f947","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"3ef86a5e9412f29d2f094beb2bb91fb1e9cb1bfb9566edce49a15382b527f947","first_computed_at":"2026-07-05T10:03:32.024780Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T10:03:32.024780Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"v9Dz1q9E28z4CDdMpF5S9yZRjm7T3qvRuWYp/yepmO3eUt0R6NGZBFjgzgJCy0D0dHSiOxn6rE25d9yeG+gSBQ==","signature_status":"signed_v1","signed_at":"2026-07-05T10:03:32.025188Z","signed_message":"canonical_sha256_bytes"},"source_id":"2501.12147","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:7a5f677aac2d1c7aa3c9eeedbe3ff880a46be180d34f23c849b3a31733025fb0","sha256:26e676c8cb847f25eb82594631fb65028992dfc461a45a789a14831e13188912"],"state_sha256":"cb1b48a38577c877d48cfa0566786bd5acd153b7a03b7a83964d483c35545b27"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"kFKvpxiD0PRTb94ttvBE8LRwUe8+raqf50lwcDaKVf/VHJ24nLA4/OH/8pTjX+gP2EZL5vNprG1Zp9CE1XTDDQ==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-10T03:14:51.029396Z","bundle_sha256":"d5145ddd8803d1f2aa723a3dbbd4f5e231376536b9d7041e79c1c220dd81c08d"}}