{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2023:LUS63OLPNFXPQTK4SMY27NKYWC","short_pith_number":"pith:LUS63OLP","canonical_record":{"source":{"id":"2310.04680","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2023-10-07T03:36:39Z","cross_cats_sorted":["cs.AI","cs.LG"],"title_canon_sha256":"536fa4a41dd3cfeacbdeb9ba01df8ca4cbceac6734a9a0bd56c7a1c4ec2a3158","abstract_canon_sha256":"c3a006dec9081fb300820d89f645ada8d0667e61ac6a1ddcc52e25032ec1bbff"},"schema_version":"1.0"},"canonical_sha256":"5d25edb96f696ef84d5c9331afb558b0a487ff7af6ac67f788dcc444a9a367de","source":{"kind":"arxiv","id":"2310.04680","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2310.04680","created_at":"2026-06-10T01:09:10Z"},{"alias_kind":"arxiv_version","alias_value":"2310.04680v1","created_at":"2026-06-10T01:09:10Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2310.04680","created_at":"2026-06-10T01:09:10Z"},{"alias_kind":"pith_short_12","alias_value":"LUS63OLPNFXP","created_at":"2026-06-10T01:09:10Z"},{"alias_kind":"pith_short_16","alias_value":"LUS63OLPNFXPQTK4","created_at":"2026-06-10T01:09:10Z"},{"alias_kind":"pith_short_8","alias_value":"LUS63OLP","created_at":"2026-06-10T01:09:10Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2023:LUS63OLPNFXPQTK4SMY27NKYWC","target":"record","payload":{"canonical_record":{"source":{"id":"2310.04680","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2023-10-07T03:36:39Z","cross_cats_sorted":["cs.AI","cs.LG"],"title_canon_sha256":"536fa4a41dd3cfeacbdeb9ba01df8ca4cbceac6734a9a0bd56c7a1c4ec2a3158","abstract_canon_sha256":"c3a006dec9081fb300820d89f645ada8d0667e61ac6a1ddcc52e25032ec1bbff"},"schema_version":"1.0"},"canonical_sha256":"5d25edb96f696ef84d5c9331afb558b0a487ff7af6ac67f788dcc444a9a367de","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-10T01:09:10.839380Z","signature_b64":"c7+FZr1wkJU9XoDCmjhBTXxGnnQfzj7zeVccN0UBqYvORVIR1rhfnd6u1Oo3jvZEWNCD3GfWcVjjCPe3ayowBg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"5d25edb96f696ef84d5c9331afb558b0a487ff7af6ac67f788dcc444a9a367de","last_reissued_at":"2026-06-10T01:09:10.838380Z","signature_status":"signed_v1","first_computed_at":"2026-06-10T01:09:10.838380Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2310.04680","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":"2026-06-10T01:09:10Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"ES0SW9qDQI5tWNVm/QkO5XvUBx3GQRLnqbI2mLUuemMEpmPCTtEr8RR/TrbF3WghEptyAzAER4vGfB3zsWAOAQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-06-28T16:49:11.711233Z"},"content_sha256":"a67653f9cdb9fdad942fddfb3943b5263267de8f0a0df4c588f004788a4586e9","schema_version":"1.0","event_id":"sha256:a67653f9cdb9fdad942fddfb3943b5263267de8f0a0df4c588f004788a4586e9"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2023:LUS63OLPNFXPQTK4SMY27NKYWC","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"The Cost of Down-Scaling Language Models: Fact Recall Deteriorates before In-Context Learning","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.LG"],"primary_cat":"cs.CL","authors_text":"Gintare Karolina Dziugaite, Jonathan Ragan-Kelley, Michael Carbin, Nolan Clement, Tian Jin, Vaishnavh Nagarajan, Xin Dong","submitted_at":"2023-10-07T03:36:39Z","abstract_excerpt":"How does scaling the number of parameters in large language models (LLMs) affect their core capabilities? We study two natural scaling techniques -- weight pruning and simply training a smaller or larger model, which we refer to as dense scaling -- and their effects on two core capabilities of LLMs: (a) recalling facts presented during pre-training and (b) processing information presented in-context during inference. By curating a suite of tasks that help disentangle these two capabilities, we find a striking difference in how these two abilities evolve due to scaling. Reducing the model size "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2310.04680","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/2310.04680/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":"2026-06-10T01:09:10Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"iz/RZJAG65YklHJ+Bbi/fIkASQrJe35hTuJhFB60o2kGStDuwJf2jb1x7rvnyKi8vaQT4NQYbErlFGvprdnZCg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-06-28T16:49:11.711650Z"},"content_sha256":"c113ab8069d7257d455a8cc237bb8ab3a4dfc662040eccb2e0015e80a3158389","schema_version":"1.0","event_id":"sha256:c113ab8069d7257d455a8cc237bb8ab3a4dfc662040eccb2e0015e80a3158389"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/LUS63OLPNFXPQTK4SMY27NKYWC/bundle.json","state_url":"https://pith.science/pith/LUS63OLPNFXPQTK4SMY27NKYWC/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/LUS63OLPNFXPQTK4SMY27NKYWC/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":"2026-06-28T16:49:11Z","links":{"resolver":"https://pith.science/pith/LUS63OLPNFXPQTK4SMY27NKYWC","bundle":"https://pith.science/pith/LUS63OLPNFXPQTK4SMY27NKYWC/bundle.json","state":"https://pith.science/pith/LUS63OLPNFXPQTK4SMY27NKYWC/state.json","well_known_bundle":"https://pith.science/.well-known/pith/LUS63OLPNFXPQTK4SMY27NKYWC/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2023:LUS63OLPNFXPQTK4SMY27NKYWC","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":"c3a006dec9081fb300820d89f645ada8d0667e61ac6a1ddcc52e25032ec1bbff","cross_cats_sorted":["cs.AI","cs.LG"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2023-10-07T03:36:39Z","title_canon_sha256":"536fa4a41dd3cfeacbdeb9ba01df8ca4cbceac6734a9a0bd56c7a1c4ec2a3158"},"schema_version":"1.0","source":{"id":"2310.04680","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2310.04680","created_at":"2026-06-10T01:09:10Z"},{"alias_kind":"arxiv_version","alias_value":"2310.04680v1","created_at":"2026-06-10T01:09:10Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2310.04680","created_at":"2026-06-10T01:09:10Z"},{"alias_kind":"pith_short_12","alias_value":"LUS63OLPNFXP","created_at":"2026-06-10T01:09:10Z"},{"alias_kind":"pith_short_16","alias_value":"LUS63OLPNFXPQTK4","created_at":"2026-06-10T01:09:10Z"},{"alias_kind":"pith_short_8","alias_value":"LUS63OLP","created_at":"2026-06-10T01:09:10Z"}],"graph_snapshots":[{"event_id":"sha256:c113ab8069d7257d455a8cc237bb8ab3a4dfc662040eccb2e0015e80a3158389","target":"graph","created_at":"2026-06-10T01:09:10Z","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/2310.04680/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"How does scaling the number of parameters in large language models (LLMs) affect their core capabilities? We study two natural scaling techniques -- weight pruning and simply training a smaller or larger model, which we refer to as dense scaling -- and their effects on two core capabilities of LLMs: (a) recalling facts presented during pre-training and (b) processing information presented in-context during inference. By curating a suite of tasks that help disentangle these two capabilities, we find a striking difference in how these two abilities evolve due to scaling. Reducing the model size ","authors_text":"Gintare Karolina Dziugaite, Jonathan Ragan-Kelley, Michael Carbin, Nolan Clement, Tian Jin, Vaishnavh Nagarajan, Xin Dong","cross_cats":["cs.AI","cs.LG"],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2023-10-07T03:36:39Z","title":"The Cost of Down-Scaling Language Models: Fact Recall Deteriorates before In-Context Learning"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2310.04680","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:a67653f9cdb9fdad942fddfb3943b5263267de8f0a0df4c588f004788a4586e9","target":"record","created_at":"2026-06-10T01:09:10Z","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":"c3a006dec9081fb300820d89f645ada8d0667e61ac6a1ddcc52e25032ec1bbff","cross_cats_sorted":["cs.AI","cs.LG"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2023-10-07T03:36:39Z","title_canon_sha256":"536fa4a41dd3cfeacbdeb9ba01df8ca4cbceac6734a9a0bd56c7a1c4ec2a3158"},"schema_version":"1.0","source":{"id":"2310.04680","kind":"arxiv","version":1}},"canonical_sha256":"5d25edb96f696ef84d5c9331afb558b0a487ff7af6ac67f788dcc444a9a367de","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"5d25edb96f696ef84d5c9331afb558b0a487ff7af6ac67f788dcc444a9a367de","first_computed_at":"2026-06-10T01:09:10.838380Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-06-10T01:09:10.838380Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"c7+FZr1wkJU9XoDCmjhBTXxGnnQfzj7zeVccN0UBqYvORVIR1rhfnd6u1Oo3jvZEWNCD3GfWcVjjCPe3ayowBg==","signature_status":"signed_v1","signed_at":"2026-06-10T01:09:10.839380Z","signed_message":"canonical_sha256_bytes"},"source_id":"2310.04680","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:a67653f9cdb9fdad942fddfb3943b5263267de8f0a0df4c588f004788a4586e9","sha256:c113ab8069d7257d455a8cc237bb8ab3a4dfc662040eccb2e0015e80a3158389"],"state_sha256":"1acf357b8d7ca4eea2823e166937e73947ee640f085701c4282f3047c2dc113d"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"NAXh9frsRXbip9uYxpa9pBoRMMbVPMcKyRe1BbI2cnGfXJ7SVha6getG5miDOdSVRiZk5nuX5L5Djpq7HIkeCQ==","signed_message":"bundle_sha256_bytes","signed_at":"2026-06-28T16:49:11.713864Z","bundle_sha256":"9ac25bf58158dfeb53707f758765aa974bf2c28817e3e4a9f82271a184831c51"}}