Community directory previewDeepSeek Harness is in developer preview
LA

llm-as-a-Verifier-dsh

PluginConfirmed pluginCommunity

aispin-dev/llm-as-a-Verifier-dsh · updated 7 days ago

Installation

dsh plugin --profile <your-profile> add @aispin/plugin-verifier
dsh plugin --profile <your-profile> add /path/to/llm-as-a-Verifier-dsh
dsh plugin --profile web add github:aispin-dev/llm-as-a-Verifier-dsh

Plugins in this repository

  • @aispin/plugin-verifierv0.3.2

    LLM-as-a-Verifier for dsh with the Best-of-N conversation mode built in: rank N candidates with a fine-grained verifier (expected grade over the logprob distribution), on demand via the verify tool or automatically on every turn of a Best-of-N session. Independent implementation of arXiv:2607.05391. By Aispin

    package.json
    package.json
    dsh.bundle.patch
    cordis.patch.yml
    Official core packages (9)
    • @deepseek-ai/cordis4.0.1shared
    • @deepseek-ai/dsh-client-runtime^0.1.0-rc.7shared
    • @deepseek-ai/dsh-client-ui-settings^0.1.0-rc.7shared
    • @deepseek-ai/dsh-client-web-react^0.1.0-rc.7shared
    • @deepseek-ai/dsh-credentials^0.1.0-rc.7shared
    • @deepseek-ai/dsh-llm^0.1.0-rc.6shared
    • @deepseek-ai/dsh-settings^0.1.0-rc.7shared
    • @deepseek-ai/dsh-tools^0.1.0-rc.6shared
    • @deepseek-ai/schemastery^3.18.1shared

Paths are pinned to commit f7fbb062dc00

Overview

LLM-as-a-Verifier (arXiv:2607.05391) as a dsh plugin — Best-of-N conversation mode: give DeepSeek V4 Flash test-time scaling. Bo5 self-verification hits 88% on Terminal-Bench 2.1, beating some frontier models at a fraction of the cost. Fine-grained logprob-expectation scoring, PPT tournament, zero-config.

Repository check result

Results are bound to the listed commit and cover only static checks run by this registry. The Evidence Score measures reviewable information; it is not a compatibility percentage or a security guarantee.

Static repository check

No static findings
Checked commit
f7fbb062dc00
Evidence Score
83/100
Static risk
No static findings
Files scanned
2
Findings
0

Evidence breakdown

Source provenance25/25
Install documentation25/25
Maintenance12/20
Compatibility evidence11/20
Security signals10/10

Compatibility confidence: Limited

Findings (0)

No static findings were recorded for this commit.

Not run
  • Dependency advisory database scan
  • Installation execution test
  • Runtime behavior verification

Scanner 1.0.0 · Scorer 1.0.0