{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2026:ASQSBGW2TJNHKVXQJQRRMQ6LVY","short_pith_number":"pith:ASQSBGW2","canonical_record":{"source":{"id":"2607.16850","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2026-07-18T15:13:33Z","cross_cats_sorted":[],"title_canon_sha256":"013343d05a7bdc4e2786de0552375c289ef68b9259179e85b827cdf95b61797e","abstract_canon_sha256":"31e0b23eba8b75456999f0f8ffc3864710bb85299b7de49e6f054e8f728eff23"},"schema_version":"1.0"},"canonical_sha256":"04a1209ada9a5a7556f04c231643cbae3b6ddef59679fae79dd39ce25a6f675a","source":{"kind":"arxiv","id":"2607.16850","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2607.16850","created_at":"2026-07-21T01:21:02Z"},{"alias_kind":"arxiv_version","alias_value":"2607.16850v1","created_at":"2026-07-21T01:21:02Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2607.16850","created_at":"2026-07-21T01:21:02Z"},{"alias_kind":"pith_short_12","alias_value":"ASQSBGW2TJNH","created_at":"2026-07-21T01:21:02Z"},{"alias_kind":"pith_short_16","alias_value":"ASQSBGW2TJNHKVXQ","created_at":"2026-07-21T01:21:02Z"},{"alias_kind":"pith_short_8","alias_value":"ASQSBGW2","created_at":"2026-07-21T01:21:02Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2026:ASQSBGW2TJNHKVXQJQRRMQ6LVY","target":"record","payload":{"canonical_record":{"source":{"id":"2607.16850","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2026-07-18T15:13:33Z","cross_cats_sorted":[],"title_canon_sha256":"013343d05a7bdc4e2786de0552375c289ef68b9259179e85b827cdf95b61797e","abstract_canon_sha256":"31e0b23eba8b75456999f0f8ffc3864710bb85299b7de49e6f054e8f728eff23"},"schema_version":"1.0"},"canonical_sha256":"04a1209ada9a5a7556f04c231643cbae3b6ddef59679fae79dd39ce25a6f675a","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-21T01:21:02.179850Z","signature_b64":"hGDvyWA3941VOb2lWARGDOqUl+tvyzxsCTotP89IdalL8GpMx7QCvOdsc0bSS9GO+YXn6jqt5es5hJmkZIXsBQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"04a1209ada9a5a7556f04c231643cbae3b6ddef59679fae79dd39ce25a6f675a","last_reissued_at":"2026-07-21T01:21:02.179004Z","signature_status":"signed_v1","first_computed_at":"2026-07-21T01:21:02.179004Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2607.16850","source_version":1,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-21T01:21:02Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"QSOHQroQd8E3/t6hqmnrFV8Fiy5i8MAUBJvs8bZiLreqYfnDewjBj49Hbs5p7iGaOyxedl8Xx6phx4LA6CtqCw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-04T16:17:43.220870Z"},"content_sha256":"755ed9641ae3edbdd317a57b2199ce2ebac915872719e2884fd1adcc2f193410","schema_version":"1.0","event_id":"sha256:755ed9641ae3edbdd317a57b2199ce2ebac915872719e2884fd1adcc2f193410"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2026:ASQSBGW2TJNHKVXQJQRRMQ6LVY","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Group Entropy-Controlled Policy Optimization","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Chengqi Lyu, Guangran Cheng, Kai Chen, Songyang Gao, Wenwei Zhang","submitted_at":"2026-07-18T15:13:33Z","abstract_excerpt":"Entropy control has become an effective tool in reinforcement learning (RL) of large language models (LLMs), helping balance exploration-exploitation trade-off during alignment process. Such RL paradigm is often conducted on mixtures of heterogeneous tasks, which induce distinct entropy regimes under the same policy, making global or token-level entropy regulation insufficient to corresponding heterogeneous needs of exploration. This heterogeneity further makes GRPO-style normalized advantages induce an entropy-dependent bias, making advantage signals across prompt groups statistically non-com"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2607.16850","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/2607.16850/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":"2026-07-21T01:21:02Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"oBUb55kDXS3hXVzTwO5DckKZtIiOjDsR4IaChsmPLYrCWw5JZ5ZM36WphdRqgUourSheeVcwn2R4ES/z4wb+AQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-04T16:17:43.221774Z"},"content_sha256":"76cfd1e82fcf0508ec5d58a5f24f6d59cc39a40421dc23c7ba4fc33c08365d86","schema_version":"1.0","event_id":"sha256:76cfd1e82fcf0508ec5d58a5f24f6d59cc39a40421dc23c7ba4fc33c08365d86"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/ASQSBGW2TJNHKVXQJQRRMQ6LVY/bundle.json","state_url":"https://pith.science/pith/ASQSBGW2TJNHKVXQJQRRMQ6LVY/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/ASQSBGW2TJNHKVXQJQRRMQ6LVY/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":"2026-08-04T16:17:43Z","links":{"resolver":"https://pith.science/pith/ASQSBGW2TJNHKVXQJQRRMQ6LVY","bundle":"https://pith.science/pith/ASQSBGW2TJNHKVXQJQRRMQ6LVY/bundle.json","state":"https://pith.science/pith/ASQSBGW2TJNHKVXQJQRRMQ6LVY/state.json","well_known_bundle":"https://pith.science/.well-known/pith/ASQSBGW2TJNHKVXQJQRRMQ6LVY/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:ASQSBGW2TJNHKVXQJQRRMQ6LVY","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":"31e0b23eba8b75456999f0f8ffc3864710bb85299b7de49e6f054e8f728eff23","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2026-07-18T15:13:33Z","title_canon_sha256":"013343d05a7bdc4e2786de0552375c289ef68b9259179e85b827cdf95b61797e"},"schema_version":"1.0","source":{"id":"2607.16850","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2607.16850","created_at":"2026-07-21T01:21:02Z"},{"alias_kind":"arxiv_version","alias_value":"2607.16850v1","created_at":"2026-07-21T01:21:02Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2607.16850","created_at":"2026-07-21T01:21:02Z"},{"alias_kind":"pith_short_12","alias_value":"ASQSBGW2TJNH","created_at":"2026-07-21T01:21:02Z"},{"alias_kind":"pith_short_16","alias_value":"ASQSBGW2TJNHKVXQ","created_at":"2026-07-21T01:21:02Z"},{"alias_kind":"pith_short_8","alias_value":"ASQSBGW2","created_at":"2026-07-21T01:21:02Z"}],"graph_snapshots":[{"event_id":"sha256:76cfd1e82fcf0508ec5d58a5f24f6d59cc39a40421dc23c7ba4fc33c08365d86","target":"graph","created_at":"2026-07-21T01:21:02Z","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/2607.16850/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Entropy control has become an effective tool in reinforcement learning (RL) of large language models (LLMs), helping balance exploration-exploitation trade-off during alignment process. Such RL paradigm is often conducted on mixtures of heterogeneous tasks, which induce distinct entropy regimes under the same policy, making global or token-level entropy regulation insufficient to corresponding heterogeneous needs of exploration. This heterogeneity further makes GRPO-style normalized advantages induce an entropy-dependent bias, making advantage signals across prompt groups statistically non-com","authors_text":"Chengqi Lyu, Guangran Cheng, Kai Chen, Songyang Gao, Wenwei Zhang","cross_cats":[],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2026-07-18T15:13:33Z","title":"Group Entropy-Controlled Policy Optimization"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2607.16850","kind":"arxiv","version":1},"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":"sha256:755ed9641ae3edbdd317a57b2199ce2ebac915872719e2884fd1adcc2f193410","target":"record","created_at":"2026-07-21T01:21:02Z","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":"31e0b23eba8b75456999f0f8ffc3864710bb85299b7de49e6f054e8f728eff23","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2026-07-18T15:13:33Z","title_canon_sha256":"013343d05a7bdc4e2786de0552375c289ef68b9259179e85b827cdf95b61797e"},"schema_version":"1.0","source":{"id":"2607.16850","kind":"arxiv","version":1}},"canonical_sha256":"04a1209ada9a5a7556f04c231643cbae3b6ddef59679fae79dd39ce25a6f675a","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"04a1209ada9a5a7556f04c231643cbae3b6ddef59679fae79dd39ce25a6f675a","first_computed_at":"2026-07-21T01:21:02.179004Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-21T01:21:02.179004Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"hGDvyWA3941VOb2lWARGDOqUl+tvyzxsCTotP89IdalL8GpMx7QCvOdsc0bSS9GO+YXn6jqt5es5hJmkZIXsBQ==","signature_status":"signed_v1","signed_at":"2026-07-21T01:21:02.179850Z","signed_message":"canonical_sha256_bytes"},"source_id":"2607.16850","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:755ed9641ae3edbdd317a57b2199ce2ebac915872719e2884fd1adcc2f193410","sha256:76cfd1e82fcf0508ec5d58a5f24f6d59cc39a40421dc23c7ba4fc33c08365d86"],"state_sha256":"d2ec7364e0e6888ff233900bf7167d0c9ef4ae2e5b9f16f9629d8c575d157bb7"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"w7cT5p/hu2EtVlvDF9lTHJs+2Ie7i+HsuiG1dUEix5LKeA9okjDOkGbV2Sg3UpVAzQU8cE7m/lpLgFKOyd9xDA==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-04T16:17:43.228829Z","bundle_sha256":"82f0e8a632515476753e0f841dba84800cfb325ead39d76056f7340c444f1cfe"}}