{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:I4PPOL3GMLXSBMHAS2WPRSLAYB","short_pith_number":"pith:I4PPOL3G","schema_version":"1.0","canonical_sha256":"471ef72f6662ef20b0e096acf8c960c05a37b3050684b9f6ebdcd5366d4602ae","source":{"kind":"arxiv","id":"2606.03618","version":1},"attestation_state":"computed","paper":{"title":"Cross-Lingual Token Arbitrage: Optimizing Code Agent Context Windows via Local LLM Preprocessing","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.AI","authors_text":"Mehmet Utku Colak","submitted_at":"2026-06-02T13:17:45Z","abstract_excerpt":"AI-assisted coding agents are bottlenecked by input-token cost. Two pathologies of raw human input drive much of this overhead: tokenization inefficiency for non-English text and structural entropy in conversational prompts. Existing approaches act reactively by compressing already-bloated contexts or intervening after failures occur.\n  We introduce a pre-flight, edge-side prompt-rewriting middleware that operates between the developer and the cloud agent. A local Llama 3.2 (3B) model performs cross-lingual translation into English, structural rewriting into a compact task-oriented format, and"},"verification_status":{"content_addressed":true,"pith_receipt":true,"author_attested":false,"weak_author_claims":0,"strong_author_claims":0,"externally_anchored":false,"storage_verified":false,"citation_signatures":0,"replication_records":0,"graph_snapshot":true,"references_resolved":false,"formal_links_present":false},"canonical_record":{"source":{"id":"2606.03618","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2026-06-02T13:17:45Z","cross_cats_sorted":[],"title_canon_sha256":"6957a30c79d288c5824445d7dde68dd33133ffd7235be7a2cb720a1cb8c03e82","abstract_canon_sha256":"9dcf75f3ebe519e2e46eb04570168aa9ac1823e5409f20aab37cd65e65526514"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-03T01:06:02.485123Z","signature_b64":"jexNdP14HlEy3VT6VMvK2own/qTEaPu8J1U6A8HPlJ7QmEbtkpLJ++PRrvu+mZTzp8KqwBkHKYjRJOENG55IDQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"471ef72f6662ef20b0e096acf8c960c05a37b3050684b9f6ebdcd5366d4602ae","last_reissued_at":"2026-06-03T01:06:02.484687Z","signature_status":"signed_v1","first_computed_at":"2026-06-03T01:06:02.484687Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Cross-Lingual Token Arbitrage: Optimizing Code Agent Context Windows via Local LLM Preprocessing","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.AI","authors_text":"Mehmet Utku Colak","submitted_at":"2026-06-02T13:17:45Z","abstract_excerpt":"AI-assisted coding agents are bottlenecked by input-token cost. Two pathologies of raw human input drive much of this overhead: tokenization inefficiency for non-English text and structural entropy in conversational prompts. Existing approaches act reactively by compressing already-bloated contexts or intervening after failures occur.\n  We introduce a pre-flight, edge-side prompt-rewriting middleware that operates between the developer and the cloud agent. A local Llama 3.2 (3B) model performs cross-lingual translation into English, structural rewriting into a compact task-oriented format, and"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2606.03618","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/2606.03618/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"},"aliases":[{"alias_kind":"arxiv","alias_value":"2606.03618","created_at":"2026-06-03T01:06:02.484772+00:00"},{"alias_kind":"arxiv_version","alias_value":"2606.03618v1","created_at":"2026-06-03T01:06:02.484772+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2606.03618","created_at":"2026-06-03T01:06:02.484772+00:00"},{"alias_kind":"pith_short_12","alias_value":"I4PPOL3GMLXS","created_at":"2026-06-03T01:06:02.484772+00:00"},{"alias_kind":"pith_short_16","alias_value":"I4PPOL3GMLXSBMHA","created_at":"2026-06-03T01:06:02.484772+00:00"},{"alias_kind":"pith_short_8","alias_value":"I4PPOL3G","created_at":"2026-06-03T01:06:02.484772+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":0,"internal_anchor_count":0,"sample":[]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/I4PPOL3GMLXSBMHAS2WPRSLAYB","json":"https://pith.science/pith/I4PPOL3GMLXSBMHAS2WPRSLAYB.json","graph_json":"https://pith.science/api/pith-number/I4PPOL3GMLXSBMHAS2WPRSLAYB/graph.json","events_json":"https://pith.science/api/pith-number/I4PPOL3GMLXSBMHAS2WPRSLAYB/events.json","paper":"https://pith.science/paper/I4PPOL3G"},"agent_actions":{"view_html":"https://pith.science/pith/I4PPOL3GMLXSBMHAS2WPRSLAYB","download_json":"https://pith.science/pith/I4PPOL3GMLXSBMHAS2WPRSLAYB.json","view_paper":"https://pith.science/paper/I4PPOL3G","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2606.03618&json=true","fetch_graph":"https://pith.science/api/pith-number/I4PPOL3GMLXSBMHAS2WPRSLAYB/graph.json","fetch_events":"https://pith.science/api/pith-number/I4PPOL3GMLXSBMHAS2WPRSLAYB/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/I4PPOL3GMLXSBMHAS2WPRSLAYB/action/timestamp_anchor","attest_storage":"https://pith.science/pith/I4PPOL3GMLXSBMHAS2WPRSLAYB/action/storage_attestation","attest_author":"https://pith.science/pith/I4PPOL3GMLXSBMHAS2WPRSLAYB/action/author_attestation","sign_citation":"https://pith.science/pith/I4PPOL3GMLXSBMHAS2WPRSLAYB/action/citation_signature","submit_replication":"https://pith.science/pith/I4PPOL3GMLXSBMHAS2WPRSLAYB/action/replication_record"}},"created_at":"2026-06-03T01:06:02.484772+00:00","updated_at":"2026-06-03T01:06:02.484772+00:00"}