skills.rest
duypham2801
duypham2801COMMUNITY
Explore all skills by this author →
GitHub

analyze-results

Parse JSON/CSV ML experiment results into comparison tables and statistics.

Updated May 25, 2026
Search Tags:#statistics#reproducibility#ml experiments#result analysis#comparison tables#hyperparameter sweeps#research insights
One-click install
npx skills add https://github.com/duypham2801/ThS_LLM --skill analyze-results-duypham2801
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: analyze-results
Source: https://github.com/duypham2801/ThS_LLM/tree/main/.claude/skills/analyze-results
Command: npx skills add https://github.com/duypham2801/ThS_LLM --skill analyze-results-duypham2801

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

It helps you quickly understand what your ML experiments actually achieved by turning scattered result files into readable comparisons, statistics, and research-ready conclusions.

Core Features & Use Cases

  • Comparison tables from result files: Automatically groups metrics by model variants, hyperparameters, and data configurations, and computes delta vs a baseline.
  • Reproducibility-aware statistics: Summarizes multi-seed behavior with mean ± standard deviation and highlights outliers or unstable runs.
  • Actionable research interpretation: Converts numeric findings into structured statements (observation, interpretation, implication, next step) to guide follow-up experiments.

Quick Start

Use analyze-results on your experiment output folder to produce a raw results table, key numbered findings, and suggested next experiments.

Frequently Asked Questions about analyze-results

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I analyze ML experiment results from multiple seeds and hyperparameter sweeps?▼

To analyze ML experiment results, this Skill parses JSON and CSV files to organize variables, compute relative improvements against baselines, and summarize mean and standard deviation across multiple seeds.

What is the best way to compare ML model performance using JSON or CSV result files?▼

Comparing ML model performance is done by automatically grouping metrics by model variants and hyperparameter configurations, generating structured comparison tables with computed baseline deltas.

How do I calculate reproducibility statistics for machine learning training outputs?▼

Reproducibility statistics are calculated by summarizing multi-seed behavior with mean and standard deviation, while automatically highlighting outliers or unstable runs within your training outputs.

Can I turn raw evaluation metrics into actionable research insights automatically?▼

Yes, raw evaluation metrics are converted into actionable research interpretation by structuring numeric findings into observation, interpretation, implication, and next step statements to guide follow-up experiments.

Does this result analysis approach work for both training and evaluation output files?▼

Yes, this result analysis approach applies to workflows involving both training and evaluation outputs, automatically parsing the files to identify independent and dependent variables for comparison.

Related Skills

Explore More Data & Analytics Skills→
dogekikidogekiki

analyze-results

Parse JSON and CSV experiment files to compute performance statistics and comparative metrics.

Community
Intermediate
MACJONSWMACJONSW

analyze-results

Analyze machine learning experiment results and generate comparison tables.

Community
Intermediate
nishide-devnishide-dev

ml-wandb-tracking

Log metrics, hyperparameters, artifacts, and checkpoints to W&B for PyTorch Lightning runs.

Community
Advanced
raja21068raja21068
2

analyze-results

Analyze ML experiment JSON/CSV outputs into comparison tables and statistical summaries.

Community
Intermediate