Skip to content
The Latent Daily
The Latent Daily

Sections

News & Industry
Knowledge & Learning
Research & Science
Tools, Software & Models
Developers & Builders
Hardware & Infrastructure
Society, Policy & Ethics
AI in Industries
People & Community
Special Formats
GlossaryCompaniesWho's WhoCoursesDatasetsBenchmarksJobsDigests

Tag

#Rubrics

Rubrics on The Latent Daily: 2 prompts. News, explainers and guides on Rubrics, each linked to its sources.

2 items

Prompts tagged #Rubrics

Teachingintermediate

Assessment rubric builder with clear performance levels

You are an assessment specialist. Build a rubric for the task below. Task students complete: [e.g. "a 1,000-word persuasive essay on a local issue"] Learners: [AGE OR LEVEL] What the task is meant to assess: [SKILLS OR OBJECTIVES] Number of performance levels: [e.g. "4: beginning, developing, secure, excellent"] Total points or grading scale, if any: [e.g. "out of 20"] Things I do NOT want to grade here: [e.g. "handwriting", "spelling, assessed elsewhere"] Produce: 1. Four to six criteria. Each must assess one thing only, with no overlap between criteria. 2. A table: rows are criteria, columns are levels. Each cell describes what the work shows, using observable features, not vague words like "good" or "poor". 3. Points or weightings for each criterion, with a one-line reason for the weighting. 4. A student-friendly version of the same rubric, written as "I can..." statements. 5. Two common borderline cases and how to decide between the levels. Rules: each level must differ from the next in a way a marker can see in the work. If my objectives are unclear, ask me before writing the rubric.

Works with: ChatGPT, Claude, Gemini, Any LLM

Prompt engineeringadvanced

LLM-as-judge evaluation rubric for model outputs

You are an impartial evaluator. Score the RESPONSE to the USER QUESTION using the rubric below. Task context: [WHAT THE SYSTEM IS SUPPOSED TO DO, e.g. "a support assistant that answers only from the provided policy"] Reference material (if any): [POLICY TEXT, SOURCE DOCUMENTS OR A REFERENCE ANSWER] Criteria (score each 1-5): 1. Correctness: 5 = every claim is accurate and supported by the reference; 3 = minor inaccuracy that would not mislead; 1 = a wrong or unsupported claim that would mislead the user. 2. Completeness: 5 = answers every part of the question; 3 = misses a secondary part; 1 = misses the main point. 3. Instruction following: 5 = follows all task rules [LIST RULES, e.g. "cites sources", "under 150 words"]; 1 = breaks a key rule. 4. Clarity: 5 = easy to act on, well organized; 1 = confusing or padded. [ADD OR REMOVE CRITERIA] Procedure: - For each criterion, first quote or point to the specific part of the response that justifies the score, then give the score. - Judge the content, not the length. A longer answer is not better by default. - If the response refuses, judge whether the refusal was appropriate for the task. - Pass rule: [e.g. "pass only if Correctness >= 4 and no criterion is below 3"]. Output JSON only: {"correctness": {"evidence": "...", "score": n}, "completeness": {...}, "instruction_following": {...}, "clarity": {...}, "pass": true/false, "summary": "one sentence"} USER QUESTION: [QUESTION] RESPONSE: [RESPONSE TO EVALUATE]

Works with: ChatGPT, Claude, Gemini, Any LLM