{
  "schema_version": "1.0.0",
  "rubric_version": "1.1",
  "license": "CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/)",
  "record": {
    "id": "I-0015",
    "slug": "gensyn-verde-repops",
    "title": "Verde and RepOps (Gensyn)",
    "aliases": [
      "Verde",
      "RepOps",
      "Reproducible Operators",
      "Gensyn Reproducible Execution Environment (REE)"
    ],
    "status": "draft",
    "last_reviewed": "2026-09-24",
    "review_interval_days": 90,
    "steward": null,
    "provenance": {
      "drafted_by": "ai",
      "reviewed_by": []
    },
    "risk_flags": [],
    "flags": [
      "provider-reported"
    ],
    "one_liner": "Gensyn's system for checking delegated machine-learning jobs, which settles disagreements between providers by re-running a single operation with bitwise-reproducible operators.",
    "summary": "Verde is a protocol from the company Gensyn for checking machine-learning jobs, such as inference, fine-tuning or training, that a client hands to untrusted compute providers. Several providers run the same job. If their results differ, a bisection game narrows the dispute to a single operation, which a referee re-runs to decide who is right. The client gets the correct result if at least one provider is honest. This needs identical results on different hardware, which Gensyn's RepOps library provides by fixing the order of floating-point operations. Gensyn states that Verde and RepOps run in production in Judge, its AI evaluation service, and publishes RepOps as closed binaries. The research paper reports that RepOps roughly doubled Llama-8B inference time, and its version covers 32-bit floating point on one GPU per provider. As of September 2026 no independent evaluation has been published.",
    "category": "cryptographic-computational",
    "secondary_categories": [],
    "verifies": [
      {
        "claim": "C-0005",
        "role": "primary",
        "note": "A client learns that a delegated inference output came from the declared model and input, if at least one provider is honest (S-1809, S-1810)."
      },
      {
        "claim": "C-0007",
        "role": "supporting",
        "note": "Also covers training and fine-tuning jobs delegated to several providers (S-1809)."
      }
    ],
    "threat_model": "adversarial",
    "adversarial_evaluation": "analysis",
    "hardware_requirement": "none",
    "prover_cooperation": "required",
    "confidentiality": "revealing",
    "depends_on": [],
    "readiness": {
      "assessment": true,
      "level": "R2",
      "rubric_version": "1.1",
      "rationale": "R2. A public runtime and a research paper demonstrate the approach. Gensyn states that it runs in its own production service, but gives no detail of that use, and no other party is documented relying on Verde.\n\n- **R1** met: the paper states the claim, that the client obtains the correct result if at least one provider is honest, together with the dispute protocol and its assumptions [[S-1809]].\n- **R2** met: Gensyn's Reproducible Execution Environment, which runs LLM inference with RepOps and writes receipts for re-execution, is public as binaries with an MIT-licensed SDK [[S-1812]]. The paper measures RepOps on NVIDIA T4, RTX 3090 and A100 GPUs, including Llama-8B inference and fine-tuning on an A100, against a stated adversary: dishonest compute providers [[S-1809]]. The paper's evaluation reports overheads. The claim of bitwise-identical results across hardware comes from Gensyn's posts [[S-1810]].\n- **R3** not met. Gensyn states that Verde and RepOps are deployed in Judge, its AI evaluation service, which it launched with a reasoning task framed as a prediction market [[S-1810]] [[S-1811]]. That is the developer's own use, and its posts do not say how Judge uses Verde or which decisions rest on it. By contrast, the developer of [[I-0001|TOPLOC]], rated R3 on its own use, published a technical report describing how that use accepts or rejects work [[S-1003]]. REE, the public release of Gensyn's runtime, is limited to reproducible LLM inference, and its operators ship as closed binaries [[S-1812]]. No other party is documented relying on Verde for a verification decision.\n- **R4** not met: no independent audit, red-team or peer-reviewed security analysis has been published.\n\nConfidence is low: a fuller account of the production use in Judge could meet R3.",
      "evidence": [
        "S-1809",
        "S-1810",
        "S-1811",
        "S-1812"
      ],
      "next_level_gaps": [
        "A documented production deployment, such as a technical account of how Judge uses Verde, or a party other than Gensyn relying on Verde for a verification decision.",
        "An independent public security evaluation of the Verde dispute protocol and of RepOps reproducibility across hardware."
      ],
      "confidence": "low",
      "assessed_by": [
        "ai-draft"
      ],
      "assessed_on": "2026-09-24",
      "status": "current",
      "dispute": null
    },
    "flaws": [],
    "blockers": [
      {
        "text": "RepOps as published supports only 32-bit floating point with one GPU per provider, and added 98% to Llama-8B inference time on an A100.",
        "theme": "performance-compatibility",
        "blocked_by": null,
        "sources": [
          "S-1809"
        ]
      },
      {
        "text": "The providers who re-run a job and the referee need the model and data, and the guarantee holds only if at least one provider is honest.",
        "theme": "privacy-leakage",
        "blocked_by": null,
        "sources": [
          "S-1809"
        ]
      }
    ],
    "challenge_themes": [
      "performance-compatibility",
      "adversarial-validation",
      "privacy-leakage",
      "protocol-soundness"
    ],
    "organizations": [],
    "people": [],
    "sources": [
      {
        "source": "S-1809",
        "supports": "refereed delegation design, guarantee and its limit, bisection, single-operator re-execution, RepOps design, overheads, FP32 and single-GPU scope, author affiliations",
        "locator": "abstract; §1; §3.2; §4 Table 2; limitations"
      },
      {
        "source": "S-1810",
        "supports": "deployment in Judge; bitwise reproducibility across hardware; tasks covered; what the guarantee does not cover (provider-reported)"
      },
      {
        "source": "S-1811",
        "supports": "Judge launch with a reasoning task framed as a prediction market (provider-reported)"
      },
      {
        "source": "S-1812",
        "supports": "public REE binaries and SDK, licences, receipts, version, current release limited to reproducible LLM inference (provider-reported)",
        "locator": "README"
      },
      {
        "source": "S-1003",
        "supports": "comparison in the rationale: TOPLOC's developer documents how its production use accepts or rejects work",
        "locator": "§2.3; §2.4.2"
      }
    ],
    "concepts": [
      "K-0008",
      "K-0009"
    ],
    "kind": "product",
    "developer": [],
    "realises": [
      "M-0002",
      "M-0006"
    ],
    "homepage": "https://www.gensyn.ai/research/verde-verification-system-in-production",
    "repo": "https://github.com/gensyn-ai/ree",
    "type": "implementation",
    "url": "https://trustbutveri.fyi/implementations/gensyn-verde-repops/",
    "source_file": "content/implementations/gensyn-verde-repops.md",
    "flags_all": [
      "provider-reported",
      "ai-drafted"
    ],
    "body_markdown": "## What it is\n\nVerde is a protocol from Gensyn for checking machine-learning programs, such as inference, fine-tuning and training of language models, that a client delegates to untrusted compute providers [[S-1809]]. It adapts refereed delegation, a cryptographic technique in which a client sends the same job to several providers and a referee settles disagreements [[S-1809]]. RepOps (Reproducible Operators) is the library that makes honest providers' results bitwise identical on different hardware [[S-1809]]. Most of the paper's authors work at Gensyn [[S-1809]].\n\nVerde relies on [[M-0002|bit-exact inference]], and for training jobs it resolves disputes over checkpoints, as in [[M-0006|training-transcript verification]]. Gensyn reports that Verde and RepOps are deployed in Judge, its service for verifiable AI evaluation [[S-1810]] [[S-1811]].\n\n## How it works\n\n1. The client gives the same job to two or more providers, and arbitration starts only if their outputs differ [[S-1810]].\n2. A two-level bisection game finds the first step, and then the first operation in the computational graph, on which the providers disagree [[S-1809]] [[S-1810]].\n3. The referee re-runs that one operation to decide which output is correct. The authors state this takes two orders of magnitude less compute than running the model [[S-1809]].\n\nRepOps fixes the order of floating-point operations for common operators such as matrix multiplication, so that an honest provider's result does not depend on its GPU [[S-1809]] [[S-1810]]. Gensyn's Reproducible Execution Environment (REE) packages this for LLM inference. It writes a receipt holding the model, prompt, output and the metadata needed to reproduce the run [[S-1812]].\n\n## Evidence\n- The paper reports RepOps overheads against PyTorch on NVIDIA T4, RTX 3090 and A100 GPUs, for DistilBERT and Llama models [[S-1809]]. For Llama-8B on an A100 with 80 GB, RepOps added 98% to inference time and 126% to LoRA fine-tuning time [[S-1809]].\n- Gensyn reports that RepOps gives bitwise-identical results across hardware, and that Verde and RepOps are deployed in Judge [[S-1810]]. Gensyn introduced Judge in August 2025 with a reasoning task framed as a prediction market [[S-1811]].\n- REE is public as an MIT-licensed SDK with proprietary binaries, and reached version 0.4.0 in June 2026 [[S-1812]].\n\n## Limitations\n- The guarantee needs at least one honest provider. The authors state that if all providers are dishonest, the referee accepts an incorrect output [[S-1809]].\n- The paper's RepOps supports 32-bit floating point only, and reproducibility only when each setup runs the program on a single GPU [[S-1809]].\n- Gensyn states that Verde shows the output was produced by the declared model and data, not that the model itself is correct [[S-1810]].\n- The RepOps kernels ship only as binaries under a proprietary licence [[S-1812]].\n- All evidence of production use comes from Gensyn [[S-1810]] [[S-1811]].",
    "body_text": "What it is Verde is a protocol from Gensyn for checking machine-learning programs, such as inference, fine-tuning and training of language models, that a client delegates to untrusted compute providers [S-1809]. It adapts refereed delegation, a cryptographic technique in which a client sends the same job to several providers and a referee settles disagreements [S-1809]. RepOps (Reproducible Operators) is the library that makes honest providers' results bitwise identical on different hardware [S-1809]. Most of the paper's authors work at Gensyn [S-1809]. Verde relies on bit-exact inference, and for training jobs it resolves disputes over checkpoints, as in training-transcript verification. Gensyn reports that Verde and RepOps are deployed in Judge, its service for verifiable AI evaluation [S-1810] [S-1811]. How it works 1. The client gives the same job to two or more providers, and arbitration starts only if their outputs differ [S-1810]. 2. A two-level bisection game finds the first step, and then the first operation in the computational graph, on which the providers disagree [S-1809] [S-1810]. 3. The referee re-runs that one operation to decide which output is correct. The authors state this takes two orders of magnitude less compute than running the model [S-1809]. RepOps fixes the order of floating-point operations for common operators such as matrix multiplication, so that an honest provider's result does not depend on its GPU [S-1809] [S-1810]. Gensyn's Reproducible Execution Environment (REE) packages this for LLM inference. It writes a receipt holding the model, prompt, output and the metadata needed to reproduce the run [S-1812]. Evidence - The paper reports RepOps overheads against PyTorch on NVIDIA T4, RTX 3090 and A100 GPUs, for DistilBERT and Llama models [S-1809]. For Llama-8B on an A100 with 80 GB, RepOps added 98% to inference time and 126% to LoRA fine-tuning time [S-1809]. - Gensyn reports that RepOps gives bitwise-identical results across hardware, and that Verde and RepOps are deployed in Judge [S-1810]. Gensyn introduced Judge in August 2025 with a reasoning task framed as a prediction market [S-1811]. - REE is public as an MIT-licensed SDK with proprietary binaries, and reached version 0.4.0 in June 2026 [S-1812]. Limitations - The guarantee needs at least one honest provider. The authors state that if all providers are dishonest, the referee accepts an incorrect output [S-1809]. - The paper's RepOps supports 32-bit floating point only, and reproducibility only when each setup runs the program on a single GPU [S-1809]. - Gensyn states that Verde shows the output was produced by the declared model and data, not that the model itself is correct [S-1810]. - The RepOps kernels ship only as binaries under a proprietary licence [S-1812]. - All evidence of production use comes from Gensyn [S-1810] [S-1811].",
    "referenced_by": [
      {
        "id": "M-0002",
        "title": "Deterministic and bit-exact inference",
        "url": "https://trustbutveri.fyi/mechanisms/deterministic-inference/"
      },
      {
        "id": "M-0012",
        "title": "Model identity attestation",
        "url": "https://trustbutveri.fyi/mechanisms/model-identity-attestation/"
      }
    ]
  }
}