{"type":"post","id":"beyond-prompts-a-systematic-study-of-llm-based-9497ba","slug":"beyond-prompts-a-systematic-study-of-llm-based-9497ba","title":"Beyond Prompts: A Systematic Study of LLM-Based Machine Translation Evaluators","description":"A Findings of EMNLP 2026 paper studies how prompt design, examples, model choice and workflow affect LLM-based machine-translation evaluation.","retrieval_nugget":"A Findings of EMNLP 2026 paper studies how prompt design, examples, model choice and workflow affect LLM-based machine-translation evaluation. Its authors introduce ORBIT, a family of evaluators that they say reaches expert-level results on selected tasks at about one-twentieth the cost of expert annotation. The work matters because it treats evaluator design as a system, not just a prompt.","published_at":"2026-11-04","updated_at":"2026-11-04","record_date":"2026-11-04","date_kind":"published_at","topics":["prompts"],"entities":["ORBIT","Association for Computational Linguistics","Findings of EMNLP 2026","Dmitry Popov","Roman Bokhyan","Ekaterina Enikeeva"],"related_patterns":["ai-engineering-evaluations-are-splitting-into-task-specific"],"novelty_label":"new_fact","what_changed":"A separate systematic study examines design choices in LLM-based machine-translation evaluation, using new expert annotations and cross-language benchmark validation.","prior_refs":["https://newruntime.com/posts/prompt-lottery-paper-paraphrasing-an-llm-as-a-6fe7a3/"],"source_url":"https://aclanthology.org/2026.findings-emnlp.2404","source_urls":["https://aclanthology.org/2026.findings-emnlp.2404"],"schema_version":"newruntime-agent-readable-v0.2","stable_id":"post:beyond-prompts-a-systematic-study-of-llm-based-9497ba","status":"published","visuals":[],"editor":{"role":"editor","name":"Andrey Reshetnikov","profile":"https://newruntime.com/owner-profile.md","profile_json":"https://newruntime.com/owner-profile.json","same_as":["https://www.linkedin.com/in/reshetnikov1/","https://t.me/lapetuse"]},"routes":{"html":"https://newruntime.com/posts/beyond-prompts-a-systematic-study-of-llm-based-9497ba/","markdown":"https://newruntime.com/posts/beyond-prompts-a-systematic-study-of-llm-based-9497ba.md","json":"https://newruntime.com/posts/beyond-prompts-a-systematic-study-of-llm-based-9497ba.json"},"markdown_url":"https://newruntime.com/posts/beyond-prompts-a-systematic-study-of-llm-based-9497ba.md","next_reads":[{"type":"topic","path":"/topics/prompts/","url":"https://newruntime.com/topics/prompts/","title":"Prompts","reason":"Explore the prompts topic hub.","media_type":"text/html","markdown_url":"https://newruntime.com/topics/prompts.md","json_url":"https://newruntime.com/topics/prompts.json"},{"type":"collection","path":"/posts/","url":"https://newruntime.com/posts/","title":"Field Notes","reason":"Open the latest editorial analysis.","media_type":"text/html","json_url":"https://newruntime.com/posts.json"},{"type":"collection","path":"/sources/","url":"https://newruntime.com/sources/","title":"Sources","reason":"See the original sources this archive cites.","media_type":"text/html"}]}