{
  "schema_version": 1,
  "generated_from_author_sources": true,
  "canonical_url": "https://aogavrilov.com/publications/inspectable-control/full-text/citations.json",
  "source_full_text_url": "https://aogavrilov.com/publications/inspectable-control/full-text/",
  "source_fragment_sha256": "0a76513a7e71c04fdfa8b456147f6c9dff29c27ecb47e2b7bd14ea077d225859",
  "section_index_url": "https://aogavrilov.com/publications/inspectable-control/full-text/sections.json",
  "section_index_sha256": "7a9ff632be03e756c84eaa5414493e3cc62e404930b0e085bca1b3021f538d72",
  "title": "Inspectable Control for Structure-Preserving Software Regeneration",
  "language": "en",
  "granularity": [
    "bibliographic-reference",
    "citation-occurrence",
    "source-block"
  ],
  "statistics": {
    "references": 5,
    "cited_references": 5,
    "citation_occurrences": 5,
    "citing_blocks": 2
  },
  "references": [
    {
      "id": "bib.bib1",
      "url": "https://aogavrilov.com/publications/inspectable-control/full-text/#bib.bib1",
      "label": "(1)",
      "citation_text": "Ali Razavi, Aaron van den Oord, and Oriol Vinyals. 2019. Generating Diverse High-Fidelity Images with VQ-VAE-2. In Advances in Neural Information Processing Systems.",
      "citation_text_sha256": "ac64261d0c6aba45c88030ba95a4c35571b3bdfb44571727272cbf6d6fb999e8",
      "occurrence_count": 1,
      "contexts": [
        {
          "block_id": "S1.p2.text",
          "block_url": "https://aogavrilov.com/publications/inspectable-control/full-text/#S1.p2.text",
          "section_id": "S1",
          "section_url": "https://aogavrilov.com/publications/inspectable-control/full-text/#S1",
          "section_heading": "1. Problem Setting and Method",
          "anchor_text": "1",
          "citation_cluster_text": "(1; 2; 3)",
          "text": "We therefore study hierarchical discrete latents as an inspectable control layer for software artifacts. Our system combines a hierarchical VQ-VAE with masked discrete regeneration (1; 2; 3). Each 64-token Python function is encoded into 16 top-level codes and 32 lower-level codes. Control is exposed through lock-and-regenerate: selected latent positions are frozen, the rest are regenerated, and the decoder reconstructs the final function. Developers therefore do not edit embeddings directly; they choose which latent positions remain fixed. We intentionally study a narrow setting—short Python functions, argmax decoding, and structural proxies—to isolate controllability before broader evaluations.",
          "text_sha256": "17d46f791c2fb0d1dafcc5faed65b6f512522ca37958d924baacc20c17e96289"
        }
      ]
    },
    {
      "id": "bib.bib2",
      "url": "https://aogavrilov.com/publications/inspectable-control/full-text/#bib.bib2",
      "label": "(2)",
      "citation_text": "Jacob Austin, Daniel D. Johnson, Jonathan Ho, Daniel Tarlow, and Rianne van den Berg. 2021. Structured Denoising Diffusion Models in Discrete State-Spaces. In Advances in Neural Information Processing Systems.",
      "citation_text_sha256": "aab625464aed60bc46b42f8e4321e4fb688853149cc2f51657ac6fd5418d1031",
      "occurrence_count": 1,
      "contexts": [
        {
          "block_id": "S1.p2.text",
          "block_url": "https://aogavrilov.com/publications/inspectable-control/full-text/#S1.p2.text",
          "section_id": "S1",
          "section_url": "https://aogavrilov.com/publications/inspectable-control/full-text/#S1",
          "section_heading": "1. Problem Setting and Method",
          "anchor_text": "2",
          "citation_cluster_text": "(1; 2; 3)",
          "text": "We therefore study hierarchical discrete latents as an inspectable control layer for software artifacts. Our system combines a hierarchical VQ-VAE with masked discrete regeneration (1; 2; 3). Each 64-token Python function is encoded into 16 top-level codes and 32 lower-level codes. Control is exposed through lock-and-regenerate: selected latent positions are frozen, the rest are regenerated, and the decoder reconstructs the final function. Developers therefore do not edit embeddings directly; they choose which latent positions remain fixed. We intentionally study a narrow setting—short Python functions, argmax decoding, and structural proxies—to isolate controllability before broader evaluations.",
          "text_sha256": "17d46f791c2fb0d1dafcc5faed65b6f512522ca37958d924baacc20c17e96289"
        }
      ]
    },
    {
      "id": "bib.bib3",
      "url": "https://aogavrilov.com/publications/inspectable-control/full-text/#bib.bib3",
      "label": "(3)",
      "citation_text": "Subham Sekhar Sahoo, Marianne Arriola, Aaron Gokaslan, Edgar Mariano Marroquin, Alexander M. Rush, Yair Schiff, Justin T. Chiu, and Volodymyr Kuleshov. 2024. Simple and Effective Masked Diffusion Language Models. In Advances in Neural Information Processing Systems.",
      "citation_text_sha256": "eb4b3d4e08935b0cd3f702b422e677ac5bf7e906424c659521e68f8b56231696",
      "occurrence_count": 1,
      "contexts": [
        {
          "block_id": "S1.p2.text",
          "block_url": "https://aogavrilov.com/publications/inspectable-control/full-text/#S1.p2.text",
          "section_id": "S1",
          "section_url": "https://aogavrilov.com/publications/inspectable-control/full-text/#S1",
          "section_heading": "1. Problem Setting and Method",
          "anchor_text": "3",
          "citation_cluster_text": "(1; 2; 3)",
          "text": "We therefore study hierarchical discrete latents as an inspectable control layer for software artifacts. Our system combines a hierarchical VQ-VAE with masked discrete regeneration (1; 2; 3). Each 64-token Python function is encoded into 16 top-level codes and 32 lower-level codes. Control is exposed through lock-and-regenerate: selected latent positions are frozen, the rest are regenerated, and the decoder reconstructs the final function. Developers therefore do not edit embeddings directly; they choose which latent positions remain fixed. We intentionally study a narrow setting—short Python functions, argmax decoding, and structural proxies—to isolate controllability before broader evaluations.",
          "text_sha256": "17d46f791c2fb0d1dafcc5faed65b6f512522ca37958d924baacc20c17e96289"
        }
      ]
    },
    {
      "id": "bib.bib4",
      "url": "https://aogavrilov.com/publications/inspectable-control/full-text/#bib.bib4",
      "label": "(4)",
      "citation_text": "Shraddha Barke, Michael B. James, and Nadia Polikarpova. 2023. Grounded Copilot: How Programmers Interact with Code-Generating Models. Proceedings of the ACM on Human-Computer Interaction 7, CSCW1 (2023), Article 78.",
      "citation_text_sha256": "01720afb361067bae46517804ad7b72689adec60994ac477528a84f7a7743842",
      "occurrence_count": 1,
      "contexts": [
        {
          "block_id": "S1.p1.text",
          "block_url": "https://aogavrilov.com/publications/inspectable-control/full-text/#S1.p1.text",
          "section_id": "S1",
          "section_url": "https://aogavrilov.com/publications/inspectable-control/full-text/#S1",
          "section_heading": "1. Problem Setting and Method",
          "anchor_text": "4",
          "citation_cluster_text": "(4; 5)",
          "text": "Many software-engineering tasks are better described as controlled modification than as free-form generation: a developer wants to preserve a high-level plan, regenerate only a bounded region, and inspect where change propagates (4; 5). Surface-token interfaces make that awkward because the directly manipulable object is text.",
          "text_sha256": "562c07d7273821808094fa5fa1ba7b42763b6eaa362edfca64576cac6783735a"
        }
      ]
    },
    {
      "id": "bib.bib5",
      "url": "https://aogavrilov.com/publications/inspectable-control/full-text/#bib.bib5",
      "label": "(5)",
      "citation_text": "Kechi Zhang, Jia Li, Ge Li, Xianjie Shi, and Zhi Jin. 2023. RepoCoder: Repository-Level Code Completion Through Iterative Retrieval and Generation. In Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing.",
      "citation_text_sha256": "5f68bddd2463167e3001827f8c547e56d5fb8efeede9d32b91b660e9e2fccd23",
      "occurrence_count": 1,
      "contexts": [
        {
          "block_id": "S1.p1.text",
          "block_url": "https://aogavrilov.com/publications/inspectable-control/full-text/#S1.p1.text",
          "section_id": "S1",
          "section_url": "https://aogavrilov.com/publications/inspectable-control/full-text/#S1",
          "section_heading": "1. Problem Setting and Method",
          "anchor_text": "5",
          "citation_cluster_text": "(4; 5)",
          "text": "Many software-engineering tasks are better described as controlled modification than as free-form generation: a developer wants to preserve a high-level plan, regenerate only a bounded region, and inspect where change propagates (4; 5). Surface-token interfaces make that awkward because the directly manipulable object is text.",
          "text_sha256": "562c07d7273821808094fa5fa1ba7b42763b6eaa362edfca64576cac6783735a"
        }
      ]
    }
  ]
}