{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:7WS6KQ6ES2XQFUALQNOZZETTTY","short_pith_number":"pith:7WS6KQ6E","schema_version":"1.0","canonical_sha256":"fda5e543c496af02d00b835d9c92739e280025ad4efd77ad48b2c9afa948829e","source":{"kind":"arxiv","id":"2607.20129","version":1},"attestation_state":"computed","paper":{"title":"CUSUM-Shaped Inference-Time Monitoring and Targeted Re-Decoding for Quantized Small Language Model Reasoning","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.AI","authors_text":"Ayoub Belfatmi, El Hassane Ettifouri, Mahaman Sanoussi Yahaya Alassan, Walid Dahhane","submitted_at":"2026-07-22T13:34:17Z","abstract_excerpt":"Quantized small autoregressive reasoning models can enter long, repetitive, or unproductive trajectories, yet inference-time compute is usually allocated without observing how a trajectory develops. Building on an earlier token-level e-CUSUM controller, we develop MGT-B (Monitoring-Guided Test-time Backtracking), a revised external controller that maps overlapping windows of pre-sampling uncertainty and degeneration features to position-conditional empirical tail probabilities, accumulates mixture betting factors with a CUSUM-shaped reset, and responds to an alarm by estimating a rollback poin"},"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":"2607.20129","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2026-07-22T13:34:17Z","cross_cats_sorted":[],"title_canon_sha256":"2fe241786af7ceb6d1f8a6c36882403df2e71fe26d72e6d72095586446589cc2","abstract_canon_sha256":"ffc1e28226f623ebf439803a1bc7f4557d4d7e77abcfb3dace560f51f38e56a8"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-23T01:25:02.915361Z","signature_b64":"UPJSar01NCJhUCisf0Z/VKbdXbJHOCoH0IjsiPKXhF7KXy9bMVE4gDyNS6wI4zQvt//GEs61dOIF75XxJfdTCw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"fda5e543c496af02d00b835d9c92739e280025ad4efd77ad48b2c9afa948829e","last_reissued_at":"2026-07-23T01:25:02.914531Z","signature_status":"signed_v1","first_computed_at":"2026-07-23T01:25:02.914531Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"CUSUM-Shaped Inference-Time Monitoring and Targeted Re-Decoding for Quantized Small Language Model Reasoning","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.AI","authors_text":"Ayoub Belfatmi, El Hassane Ettifouri, Mahaman Sanoussi Yahaya Alassan, Walid Dahhane","submitted_at":"2026-07-22T13:34:17Z","abstract_excerpt":"Quantized small autoregressive reasoning models can enter long, repetitive, or unproductive trajectories, yet inference-time compute is usually allocated without observing how a trajectory develops. Building on an earlier token-level e-CUSUM controller, we develop MGT-B (Monitoring-Guided Test-time Backtracking), a revised external controller that maps overlapping windows of pre-sampling uncertainty and degeneration features to position-conditional empirical tail probabilities, accumulates mixture betting factors with a CUSUM-shaped reset, and responds to an alarm by estimating a rollback poin"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2607.20129","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/2607.20129/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":"2607.20129","created_at":"2026-07-23T01:25:02.914950+00:00"},{"alias_kind":"arxiv_version","alias_value":"2607.20129v1","created_at":"2026-07-23T01:25:02.914950+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2607.20129","created_at":"2026-07-23T01:25:02.914950+00:00"},{"alias_kind":"pith_short_12","alias_value":"7WS6KQ6ES2XQ","created_at":"2026-07-23T01:25:02.914950+00:00"},{"alias_kind":"pith_short_16","alias_value":"7WS6KQ6ES2XQFUAL","created_at":"2026-07-23T01:25:02.914950+00:00"},{"alias_kind":"pith_short_8","alias_value":"7WS6KQ6E","created_at":"2026-07-23T01:25:02.914950+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/7WS6KQ6ES2XQFUALQNOZZETTTY","json":"https://pith.science/pith/7WS6KQ6ES2XQFUALQNOZZETTTY.json","graph_json":"https://pith.science/api/pith-number/7WS6KQ6ES2XQFUALQNOZZETTTY/graph.json","events_json":"https://pith.science/api/pith-number/7WS6KQ6ES2XQFUALQNOZZETTTY/events.json","paper":"https://pith.science/paper/7WS6KQ6E"},"agent_actions":{"view_html":"https://pith.science/pith/7WS6KQ6ES2XQFUALQNOZZETTTY","download_json":"https://pith.science/pith/7WS6KQ6ES2XQFUALQNOZZETTTY.json","view_paper":"https://pith.science/paper/7WS6KQ6E","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2607.20129&json=true","fetch_graph":"https://pith.science/api/pith-number/7WS6KQ6ES2XQFUALQNOZZETTTY/graph.json","fetch_events":"https://pith.science/api/pith-number/7WS6KQ6ES2XQFUALQNOZZETTTY/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/7WS6KQ6ES2XQFUALQNOZZETTTY/action/timestamp_anchor","attest_storage":"https://pith.science/pith/7WS6KQ6ES2XQFUALQNOZZETTTY/action/storage_attestation","attest_author":"https://pith.science/pith/7WS6KQ6ES2XQFUALQNOZZETTTY/action/author_attestation","sign_citation":"https://pith.science/pith/7WS6KQ6ES2XQFUALQNOZZETTTY/action/citation_signature","submit_replication":"https://pith.science/pith/7WS6KQ6ES2XQFUALQNOZZETTTY/action/replication_record"}},"created_at":"2026-07-23T01:25:02.914950+00:00","updated_at":"2026-07-23T01:25:02.914950+00:00"}