{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2023:WF36QSC6IERIBX3AR7GR2HZCCC","short_pith_number":"pith:WF36QSC6","canonical_record":{"source":{"id":"2305.14788","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2023-05-24T06:42:44Z","cross_cats_sorted":[],"title_canon_sha256":"22b694fa7ce8c3bc1debc919e7481e89e09c7eb127cabcf5cf61878271e11879","abstract_canon_sha256":"0bc3ac8de9aa045a08620d168b903c334b348b8f788dc5436b2cd13d8473ace2"},"schema_version":"1.0"},"canonical_sha256":"b177e8485e412280df608fcd1d1f2210baa80b042bdcfd7133d50112fcd62af8","source":{"kind":"arxiv","id":"2305.14788","version":2},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2305.14788","created_at":"2026-07-05T07:08:57Z"},{"alias_kind":"arxiv_version","alias_value":"2305.14788v2","created_at":"2026-07-05T07:08:57Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2305.14788","created_at":"2026-07-05T07:08:57Z"},{"alias_kind":"pith_short_12","alias_value":"WF36QSC6IERI","created_at":"2026-07-05T07:08:57Z"},{"alias_kind":"pith_short_16","alias_value":"WF36QSC6IERIBX3A","created_at":"2026-07-05T07:08:57Z"},{"alias_kind":"pith_short_8","alias_value":"WF36QSC6","created_at":"2026-07-05T07:08:57Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2023:WF36QSC6IERIBX3AR7GR2HZCCC","target":"record","payload":{"canonical_record":{"source":{"id":"2305.14788","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2023-05-24T06:42:44Z","cross_cats_sorted":[],"title_canon_sha256":"22b694fa7ce8c3bc1debc919e7481e89e09c7eb127cabcf5cf61878271e11879","abstract_canon_sha256":"0bc3ac8de9aa045a08620d168b903c334b348b8f788dc5436b2cd13d8473ace2"},"schema_version":"1.0"},"canonical_sha256":"b177e8485e412280df608fcd1d1f2210baa80b042bdcfd7133d50112fcd62af8","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T07:08:57.984077Z","signature_b64":"5s2Q07+aY1yK/DpdzBLvqtlQ8VWUuYd5EZrEAUkCZzSU0O8BJ4RP9HRbOck/A1yJjYUt6hnf894d+wJT1mj8DQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"b177e8485e412280df608fcd1d1f2210baa80b042bdcfd7133d50112fcd62af8","last_reissued_at":"2026-07-05T07:08:57.983602Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T07:08:57.983602Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2305.14788","source_version":2,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T07:08:57Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"IZBNSeUUGPgGCR9yoL1ext3SAUqART31vAGetjIEo7kRHjy3VVISSNt5fqcdn8hBJTT6R8CMuAE5qoGpzdk/BQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-07-11T12:15:25.147723Z"},"content_sha256":"f21fd03d2e3564ddd833c3f634ec1fd2666866e235921ae94ec8db2ab591bdcf","schema_version":"1.0","event_id":"sha256:f21fd03d2e3564ddd833c3f634ec1fd2666866e235921ae94ec8db2ab591bdcf"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2023:WF36QSC6IERIBX3AR7GR2HZCCC","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Adapting Language Models to Compress Contexts","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Alexander Wettig, Alexis Chevalier, Anirudh Ajith, Danqi Chen","submitted_at":"2023-05-24T06:42:44Z","abstract_excerpt":"Transformer-based language models (LMs) are powerful and widely-applicable tools, but their usefulness is constrained by a finite context window and the expensive computational cost of processing long text documents. We propose to adapt pre-trained LMs into AutoCompressors. These language models are capable of compressing long contexts into compact summary vectors, which are then accessible to the model as soft prompts. Summary vectors are trained with an unsupervised objective, whereby long documents are processed in segments, and summary vectors from all previous segments are used in languag"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2305.14788","kind":"arxiv","version":2},"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/2305.14788/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":"2026-07-05T07:08:57Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"9NcY340634+yYiHHtwcSHFI2b2uQ1c4ZPHNwG4IkoPysXU4Es02gM48loCbt/DXDRCK25DCATz48IuhzJN+jAw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-07-11T12:15:25.148111Z"},"content_sha256":"828e3391e9aac743257a534759715cf9c46236fee4007eed505940919a2ff00c","schema_version":"1.0","event_id":"sha256:828e3391e9aac743257a534759715cf9c46236fee4007eed505940919a2ff00c"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/WF36QSC6IERIBX3AR7GR2HZCCC/bundle.json","state_url":"https://pith.science/pith/WF36QSC6IERIBX3AR7GR2HZCCC/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/WF36QSC6IERIBX3AR7GR2HZCCC/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":"2026-07-11T12:15:25Z","links":{"resolver":"https://pith.science/pith/WF36QSC6IERIBX3AR7GR2HZCCC","bundle":"https://pith.science/pith/WF36QSC6IERIBX3AR7GR2HZCCC/bundle.json","state":"https://pith.science/pith/WF36QSC6IERIBX3AR7GR2HZCCC/state.json","well_known_bundle":"https://pith.science/.well-known/pith/WF36QSC6IERIBX3AR7GR2HZCCC/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2023:WF36QSC6IERIBX3AR7GR2HZCCC","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":"0bc3ac8de9aa045a08620d168b903c334b348b8f788dc5436b2cd13d8473ace2","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2023-05-24T06:42:44Z","title_canon_sha256":"22b694fa7ce8c3bc1debc919e7481e89e09c7eb127cabcf5cf61878271e11879"},"schema_version":"1.0","source":{"id":"2305.14788","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2305.14788","created_at":"2026-07-05T07:08:57Z"},{"alias_kind":"arxiv_version","alias_value":"2305.14788v2","created_at":"2026-07-05T07:08:57Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2305.14788","created_at":"2026-07-05T07:08:57Z"},{"alias_kind":"pith_short_12","alias_value":"WF36QSC6IERI","created_at":"2026-07-05T07:08:57Z"},{"alias_kind":"pith_short_16","alias_value":"WF36QSC6IERIBX3A","created_at":"2026-07-05T07:08:57Z"},{"alias_kind":"pith_short_8","alias_value":"WF36QSC6","created_at":"2026-07-05T07:08:57Z"}],"graph_snapshots":[{"event_id":"sha256:828e3391e9aac743257a534759715cf9c46236fee4007eed505940919a2ff00c","target":"graph","created_at":"2026-07-05T07:08:57Z","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/2305.14788/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Transformer-based language models (LMs) are powerful and widely-applicable tools, but their usefulness is constrained by a finite context window and the expensive computational cost of processing long text documents. We propose to adapt pre-trained LMs into AutoCompressors. These language models are capable of compressing long contexts into compact summary vectors, which are then accessible to the model as soft prompts. Summary vectors are trained with an unsupervised objective, whereby long documents are processed in segments, and summary vectors from all previous segments are used in languag","authors_text":"Alexander Wettig, Alexis Chevalier, Anirudh Ajith, Danqi Chen","cross_cats":[],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2023-05-24T06:42:44Z","title":"Adapting Language Models to Compress Contexts"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2305.14788","kind":"arxiv","version":2},"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:f21fd03d2e3564ddd833c3f634ec1fd2666866e235921ae94ec8db2ab591bdcf","target":"record","created_at":"2026-07-05T07:08:57Z","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":"0bc3ac8de9aa045a08620d168b903c334b348b8f788dc5436b2cd13d8473ace2","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2023-05-24T06:42:44Z","title_canon_sha256":"22b694fa7ce8c3bc1debc919e7481e89e09c7eb127cabcf5cf61878271e11879"},"schema_version":"1.0","source":{"id":"2305.14788","kind":"arxiv","version":2}},"canonical_sha256":"b177e8485e412280df608fcd1d1f2210baa80b042bdcfd7133d50112fcd62af8","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"b177e8485e412280df608fcd1d1f2210baa80b042bdcfd7133d50112fcd62af8","first_computed_at":"2026-07-05T07:08:57.983602Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T07:08:57.983602Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"5s2Q07+aY1yK/DpdzBLvqtlQ8VWUuYd5EZrEAUkCZzSU0O8BJ4RP9HRbOck/A1yJjYUt6hnf894d+wJT1mj8DQ==","signature_status":"signed_v1","signed_at":"2026-07-05T07:08:57.984077Z","signed_message":"canonical_sha256_bytes"},"source_id":"2305.14788","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:f21fd03d2e3564ddd833c3f634ec1fd2666866e235921ae94ec8db2ab591bdcf","sha256:828e3391e9aac743257a534759715cf9c46236fee4007eed505940919a2ff00c"],"state_sha256":"a4613eff9a2c70108ce6a1d4454556cf87e39a09c0f9c69daad72a177698bee7"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"/dpmXcg0xmoflfja8U2iFEZ/SvpwcC+koYr3m10EPvh59Fbu+p3LO3nJ7q1WmUguyVpFFrbwHwNnJdsca7thBQ==","signed_message":"bundle_sha256_bytes","signed_at":"2026-07-11T12:15:25.150659Z","bundle_sha256":"7e469fa139f384bc4d6eba0a410382416245a979b50a53e18799ae9deab6b316"}}