{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:M3EGOZNBQBAPQBOSIUXCNCJUQE","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":"865c80228ecfdf21a32311afcd873fb150bdff1f05948c4f7ff3f04f10c64e50","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CV","submitted_at":"2026-05-28T17:20:31Z","title_canon_sha256":"44fdf5b9fac4338896070680fe9dd5962e0514ed0bc67628dce11e0360ffc1f2"},"schema_version":"1.0","source":{"id":"2605.30257","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2605.30257","created_at":"2026-05-29T02:06:14Z"},{"alias_kind":"arxiv_version","alias_value":"2605.30257v1","created_at":"2026-05-29T02:06:14Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.30257","created_at":"2026-05-29T02:06:14Z"},{"alias_kind":"pith_short_12","alias_value":"M3EGOZNBQBAP","created_at":"2026-05-29T02:06:14Z"},{"alias_kind":"pith_short_16","alias_value":"M3EGOZNBQBAPQBOS","created_at":"2026-05-29T02:06:14Z"},{"alias_kind":"pith_short_8","alias_value":"M3EGOZNB","created_at":"2026-05-29T02:06:14Z"}],"graph_snapshots":[{"event_id":"sha256:01680e60fc83967563fc486c178a0b647d02abbb78655607ccecf9d40ba4c009","target":"graph","created_at":"2026-05-29T02:06:14Z","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/2605.30257/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"We present Stable-Layers, a reinforcement learning framework that eliminates the need for paired supervision by fine-tuning a pretrained layer decomposition model using only feedback from a vision-language model (VLM). Starting from Qwen-Image-Layered, we apply Flow-GRPO with LoRA adaptation, sampling multiple candidate decompositions per image, scoring them with a VLM, and optimising the policy from group-relative advantages. The key challenge lies in designing a reliable reward signal: VLMs scoring samples in isolation tend to compress their judgements into a narrow band, leaving GRPO with l","authors_text":"Ciara Rowles, Mark Boss, Nikhil Pinnaparaju, Reshinth Adithyan, Vikram Voleti","cross_cats":[],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CV","submitted_at":"2026-05-28T17:20:31Z","title":"Stable-Layers: Fine-Tuning Image Layer Decomposition Models with VLM-Scored Reinforcement Learning"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2605.30257","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":"sha256:1bd39b0800b702a046f4f34d7295123adee4ac4571e3c5ef4438f4a54ca2d1b8","target":"record","created_at":"2026-05-29T02:06:14Z","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":"865c80228ecfdf21a32311afcd873fb150bdff1f05948c4f7ff3f04f10c64e50","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CV","submitted_at":"2026-05-28T17:20:31Z","title_canon_sha256":"44fdf5b9fac4338896070680fe9dd5962e0514ed0bc67628dce11e0360ffc1f2"},"schema_version":"1.0","source":{"id":"2605.30257","kind":"arxiv","version":1}},"canonical_sha256":"66c86765a18040f805d2452e268934812cbb18a6a1ef6bfd947e79b877d7bd9d","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"66c86765a18040f805d2452e268934812cbb18a6a1ef6bfd947e79b877d7bd9d","first_computed_at":"2026-05-29T02:06:14.241674Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-05-29T02:06:14.241674Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"PzjAJ6k2xplS02xi1uq58pNi+e9wGSHw7zJJ5gRN8vSa6KT7pdRYGT43ti7VzAu/CyZ+WYU9mBXosK7WA3h6Bw==","signature_status":"signed_v1","signed_at":"2026-05-29T02:06:14.242099Z","signed_message":"canonical_sha256_bytes"},"source_id":"2605.30257","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:1bd39b0800b702a046f4f34d7295123adee4ac4571e3c5ef4438f4a54ca2d1b8","sha256:01680e60fc83967563fc486c178a0b647d02abbb78655607ccecf9d40ba4c009"],"state_sha256":"d4578bd321e1c4e26ae8d11adeb2e92f0f97271c23be2683c2a267efa794e5da"}