{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:N64AJMSCB2S55RMXO2TFVAPQ3D","short_pith_number":"pith:N64AJMSC","schema_version":"1.0","canonical_sha256":"6fb804b2420ea5dec59776a65a81f0d8f3008979e3e03da08a9241f9d4e965ea","source":{"kind":"arxiv","id":"2606.05296","version":1},"attestation_state":"computed","paper":{"title":"Agentic Monte Carlo: Simulating Reinforcement Learning for Black-Box Agents","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.LG","authors_text":"Anthony L. Caterini, Brendan Leigh Ross, Dae Yon Hwang, Jesse C. Cresswell, No\\\"el Vouitsis, Raunaq Suri, Valentin Villecroze","submitted_at":"2026-06-03T18:00:07Z","abstract_excerpt":"LLM agents operate in two distinct regimes: open-weight agents amenable to reinforcement learning (RL) and black-box agents whose behaviour must be controlled purely at test time. Although black-box agents are often backed by state-of-the-art proprietary LLMs, API-only access precludes parameter-level optimization, rendering most RL methods inapplicable. To address this limitation, we turn to a known equivalence between RL and Bayesian inference. We propose Agentic Monte Carlo (AMC) to directly sample from the optimal policy of a black-box agent rather than training it through RL. The optimal "},"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":"2606.05296","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-06-03T18:00:07Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"2ca072a1d58eec90375311372c4ff829f7672b056b09026958265e0e9a348eb1","abstract_canon_sha256":"ef30de79ef31780c87d1b1be61e53de427984b2ef0dfed39bbe88ad267a314b5"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-05T00:13:52.412859Z","signature_b64":"qqx5bxsPoB6pqd9EL7qTn5lzuqBfnhQjRk8EngivN4sPDebfBTfZtE9hqLwj/lrWNHM8rtniPTYLsy7jsC0dBA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"6fb804b2420ea5dec59776a65a81f0d8f3008979e3e03da08a9241f9d4e965ea","last_reissued_at":"2026-06-05T00:13:52.412358Z","signature_status":"signed_v1","first_computed_at":"2026-06-05T00:13:52.412358Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Agentic Monte Carlo: Simulating Reinforcement Learning for Black-Box Agents","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.LG","authors_text":"Anthony L. Caterini, Brendan Leigh Ross, Dae Yon Hwang, Jesse C. Cresswell, No\\\"el Vouitsis, Raunaq Suri, Valentin Villecroze","submitted_at":"2026-06-03T18:00:07Z","abstract_excerpt":"LLM agents operate in two distinct regimes: open-weight agents amenable to reinforcement learning (RL) and black-box agents whose behaviour must be controlled purely at test time. Although black-box agents are often backed by state-of-the-art proprietary LLMs, API-only access precludes parameter-level optimization, rendering most RL methods inapplicable. To address this limitation, we turn to a known equivalence between RL and Bayesian inference. We propose Agentic Monte Carlo (AMC) to directly sample from the optimal policy of a black-box agent rather than training it through RL. The optimal "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2606.05296","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/2606.05296/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":"2606.05296","created_at":"2026-06-05T00:13:52.412458+00:00"},{"alias_kind":"arxiv_version","alias_value":"2606.05296v1","created_at":"2026-06-05T00:13:52.412458+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2606.05296","created_at":"2026-06-05T00:13:52.412458+00:00"},{"alias_kind":"pith_short_12","alias_value":"N64AJMSCB2S5","created_at":"2026-06-05T00:13:52.412458+00:00"},{"alias_kind":"pith_short_16","alias_value":"N64AJMSCB2S55RMX","created_at":"2026-06-05T00:13:52.412458+00:00"},{"alias_kind":"pith_short_8","alias_value":"N64AJMSC","created_at":"2026-06-05T00:13:52.412458+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/N64AJMSCB2S55RMXO2TFVAPQ3D","json":"https://pith.science/pith/N64AJMSCB2S55RMXO2TFVAPQ3D.json","graph_json":"https://pith.science/api/pith-number/N64AJMSCB2S55RMXO2TFVAPQ3D/graph.json","events_json":"https://pith.science/api/pith-number/N64AJMSCB2S55RMXO2TFVAPQ3D/events.json","paper":"https://pith.science/paper/N64AJMSC"},"agent_actions":{"view_html":"https://pith.science/pith/N64AJMSCB2S55RMXO2TFVAPQ3D","download_json":"https://pith.science/pith/N64AJMSCB2S55RMXO2TFVAPQ3D.json","view_paper":"https://pith.science/paper/N64AJMSC","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2606.05296&json=true","fetch_graph":"https://pith.science/api/pith-number/N64AJMSCB2S55RMXO2TFVAPQ3D/graph.json","fetch_events":"https://pith.science/api/pith-number/N64AJMSCB2S55RMXO2TFVAPQ3D/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/N64AJMSCB2S55RMXO2TFVAPQ3D/action/timestamp_anchor","attest_storage":"https://pith.science/pith/N64AJMSCB2S55RMXO2TFVAPQ3D/action/storage_attestation","attest_author":"https://pith.science/pith/N64AJMSCB2S55RMXO2TFVAPQ3D/action/author_attestation","sign_citation":"https://pith.science/pith/N64AJMSCB2S55RMXO2TFVAPQ3D/action/citation_signature","submit_replication":"https://pith.science/pith/N64AJMSCB2S55RMXO2TFVAPQ3D/action/replication_record"}},"created_at":"2026-06-05T00:13:52.412458+00:00","updated_at":"2026-06-05T00:13:52.412458+00:00"}