{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2024:NVSUSLHTANHFK7WWHXFLB2M7FV","short_pith_number":"pith:NVSUSLHT","canonical_record":{"source":{"id":"2408.16090","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2024-08-28T18:44:02Z","cross_cats_sorted":[],"title_canon_sha256":"672cb75ceeee59dc3816a96c07da541c47e09c39009eeed555f4ce9f4d98b15f","abstract_canon_sha256":"41feb32f8879bfdd296eb8844c7e22605d587ddfc195bfa171bf5ad5e75cd685"},"schema_version":"1.0"},"canonical_sha256":"6d65492cf3034e557ed63dcab0e99f2d57809a5f0778c3758ec3acbd0fac0467","source":{"kind":"arxiv","id":"2408.16090","version":2},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2408.16090","created_at":"2026-07-05T09:15:37Z"},{"alias_kind":"arxiv_version","alias_value":"2408.16090v2","created_at":"2026-07-05T09:15:37Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2408.16090","created_at":"2026-07-05T09:15:37Z"},{"alias_kind":"pith_short_12","alias_value":"NVSUSLHTANHF","created_at":"2026-07-05T09:15:37Z"},{"alias_kind":"pith_short_16","alias_value":"NVSUSLHTANHFK7WW","created_at":"2026-07-05T09:15:37Z"},{"alias_kind":"pith_short_8","alias_value":"NVSUSLHT","created_at":"2026-07-05T09:15:37Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2024:NVSUSLHTANHFK7WWHXFLB2M7FV","target":"record","payload":{"canonical_record":{"source":{"id":"2408.16090","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2024-08-28T18:44:02Z","cross_cats_sorted":[],"title_canon_sha256":"672cb75ceeee59dc3816a96c07da541c47e09c39009eeed555f4ce9f4d98b15f","abstract_canon_sha256":"41feb32f8879bfdd296eb8844c7e22605d587ddfc195bfa171bf5ad5e75cd685"},"schema_version":"1.0"},"canonical_sha256":"6d65492cf3034e557ed63dcab0e99f2d57809a5f0778c3758ec3acbd0fac0467","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T09:15:37.198567Z","signature_b64":"IsUJp/2OkxbCNrAt3f9wRXHBncQ9vkOB9/XALEeGZ7Ahbdd5B89boX9mSCLUUiZagbwQu0OWTaG9XAQ/Ftj4Ag==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"6d65492cf3034e557ed63dcab0e99f2d57809a5f0778c3758ec3acbd0fac0467","last_reissued_at":"2026-07-05T09:15:37.198032Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T09:15:37.198032Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2408.16090","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":"2026-07-05T09:15:37Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"GFZNBd+1Job02xgokT9MJOolE5UsTR7heNlbdRnRSZ+QiLlyhct1JR3D/xcCKHebfuwzCOgCLoki8jbVSo0bCg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-07-31T14:03:27.260612Z"},"content_sha256":"54b59d16cc1f67cca62702c3d676e9bef4108e0131960ccee9eece9eae881d02","schema_version":"1.0","event_id":"sha256:54b59d16cc1f67cca62702c3d676e9bef4108e0131960ccee9eece9eae881d02"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2024:NVSUSLHTANHFK7WWHXFLB2M7FV","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"EPO: Hierarchical LLM Agents with Environment Preference Optimization","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Chen Sun, George Konidaris, Haotian Fu, Qi Zhao","submitted_at":"2024-08-28T18:44:02Z","abstract_excerpt":"Long-horizon decision-making tasks present significant challenges for LLM-based agents due to the need for extensive planning over multiple steps. In this paper, we propose a hierarchical framework that decomposes complex tasks into manageable subgoals, utilizing separate LLMs for subgoal prediction and low-level action generation. To address the challenge of creating training signals for unannotated datasets, we develop a reward model that leverages multimodal environment feedback to automatically generate reward signals. We introduce Environment Preference Optimization (EPO), a novel method "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2408.16090","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/2408.16090/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":"2026-07-05T09:15:37Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"hASoS+vdQ36RgTOrJ/zNwTJylVze7OsmWlAgEi9n/j7q5cJOpzGnAzE3iPOGpGb1w2OFvDEnbv3Dg2oj1vR9CA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-07-31T14:03:27.260905Z"},"content_sha256":"79efccb776609c3b8c3c4be4c0547cb69474ddce51d43dee6272794db6340590","schema_version":"1.0","event_id":"sha256:79efccb776609c3b8c3c4be4c0547cb69474ddce51d43dee6272794db6340590"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/NVSUSLHTANHFK7WWHXFLB2M7FV/bundle.json","state_url":"https://pith.science/pith/NVSUSLHTANHFK7WWHXFLB2M7FV/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/NVSUSLHTANHFK7WWHXFLB2M7FV/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":"2026-07-31T14:03:27Z","links":{"resolver":"https://pith.science/pith/NVSUSLHTANHFK7WWHXFLB2M7FV","bundle":"https://pith.science/pith/NVSUSLHTANHFK7WWHXFLB2M7FV/bundle.json","state":"https://pith.science/pith/NVSUSLHTANHFK7WWHXFLB2M7FV/state.json","well_known_bundle":"https://pith.science/.well-known/pith/NVSUSLHTANHFK7WWHXFLB2M7FV/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2024:NVSUSLHTANHFK7WWHXFLB2M7FV","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":"41feb32f8879bfdd296eb8844c7e22605d587ddfc195bfa171bf5ad5e75cd685","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2024-08-28T18:44:02Z","title_canon_sha256":"672cb75ceeee59dc3816a96c07da541c47e09c39009eeed555f4ce9f4d98b15f"},"schema_version":"1.0","source":{"id":"2408.16090","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2408.16090","created_at":"2026-07-05T09:15:37Z"},{"alias_kind":"arxiv_version","alias_value":"2408.16090v2","created_at":"2026-07-05T09:15:37Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2408.16090","created_at":"2026-07-05T09:15:37Z"},{"alias_kind":"pith_short_12","alias_value":"NVSUSLHTANHF","created_at":"2026-07-05T09:15:37Z"},{"alias_kind":"pith_short_16","alias_value":"NVSUSLHTANHFK7WW","created_at":"2026-07-05T09:15:37Z"},{"alias_kind":"pith_short_8","alias_value":"NVSUSLHT","created_at":"2026-07-05T09:15:37Z"}],"graph_snapshots":[{"event_id":"sha256:79efccb776609c3b8c3c4be4c0547cb69474ddce51d43dee6272794db6340590","target":"graph","created_at":"2026-07-05T09:15:37Z","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/2408.16090/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Long-horizon decision-making tasks present significant challenges for LLM-based agents due to the need for extensive planning over multiple steps. In this paper, we propose a hierarchical framework that decomposes complex tasks into manageable subgoals, utilizing separate LLMs for subgoal prediction and low-level action generation. To address the challenge of creating training signals for unannotated datasets, we develop a reward model that leverages multimodal environment feedback to automatically generate reward signals. We introduce Environment Preference Optimization (EPO), a novel method ","authors_text":"Chen Sun, George Konidaris, Haotian Fu, Qi Zhao","cross_cats":[],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2024-08-28T18:44:02Z","title":"EPO: Hierarchical LLM Agents with Environment Preference Optimization"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2408.16090","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":"sha256:54b59d16cc1f67cca62702c3d676e9bef4108e0131960ccee9eece9eae881d02","target":"record","created_at":"2026-07-05T09:15:37Z","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":"41feb32f8879bfdd296eb8844c7e22605d587ddfc195bfa171bf5ad5e75cd685","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2024-08-28T18:44:02Z","title_canon_sha256":"672cb75ceeee59dc3816a96c07da541c47e09c39009eeed555f4ce9f4d98b15f"},"schema_version":"1.0","source":{"id":"2408.16090","kind":"arxiv","version":2}},"canonical_sha256":"6d65492cf3034e557ed63dcab0e99f2d57809a5f0778c3758ec3acbd0fac0467","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"6d65492cf3034e557ed63dcab0e99f2d57809a5f0778c3758ec3acbd0fac0467","first_computed_at":"2026-07-05T09:15:37.198032Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T09:15:37.198032Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"IsUJp/2OkxbCNrAt3f9wRXHBncQ9vkOB9/XALEeGZ7Ahbdd5B89boX9mSCLUUiZagbwQu0OWTaG9XAQ/Ftj4Ag==","signature_status":"signed_v1","signed_at":"2026-07-05T09:15:37.198567Z","signed_message":"canonical_sha256_bytes"},"source_id":"2408.16090","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:54b59d16cc1f67cca62702c3d676e9bef4108e0131960ccee9eece9eae881d02","sha256:79efccb776609c3b8c3c4be4c0547cb69474ddce51d43dee6272794db6340590"],"state_sha256":"5b0e251b7b3daed7ccc334116538db653a207c711dde6f0bba99bbff12ad9d88"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"t0A+qawx+1r/LKULBIkxjmQmCjS7qvw4gz8T8YFkXgp0vhGyxA2kIKEtbz5Q0/V5OjFSnITln4Fvy6RCrOdwCw==","signed_message":"bundle_sha256_bytes","signed_at":"2026-07-31T14:03:27.266257Z","bundle_sha256":"fa168accd129ff8ffe65e9c4ce6f523743d5f29406ec986034e78913f026620f"}}