agentsclimarketplace

Run2 pareto optimization

Skill cxcscmu/SkillLearnBench/skills/b2-self-feedback-gemini-3.1-pro-preview/dbscan-parameter-tuning/run2_pareto-optimization

[COLM'26] SkillLearnBench is the first benchmark for evaluating continual learning methods that automatically generate agent skills.

Install
npx -y skills add cxcscmu/SkillLearnBench --skill run2_pareto-optimization

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

What its author says it does

Copied from the file, not written here

Finding the Pareto frontier for multi-objective optimization using dominance conditions.

SKILL.md

1.6 KB, 339 tokens by cl100k_base, as published. Nobody here has run it

Pareto Optimization (Dominance Condition)

When you need to optimize multiple conflicting objectives (e.g., maximizing F1, minimizing error), you can identify the Pareto frontier by explicitly checking if a point is "dominated" by another point.

Installation

Ensure you have numpy and pandas installed: pip install numpy pandas

Usage

import numpy as np
import pandas as pd

def get_pareto_frontier(df, obj_max_cols, obj_min_cols):
    """
    Finds non-dominated rows in a DataFrame.
    """
    pareto_mask = np.ones(len(df), dtype=bool)
    
    for i in range(len(df)):
        # Assume self is not dominated unless proven otherwise
        row_i = df.iloc[i]
        
        # Check against all other rows
        better_or_equal = np.ones(len(df), dtype=bool)
        strictly_better = np.zeros(len(df), dtype=bool)
        
        for col in obj_max_cols:
            better_or_equal &= (df[col] >= row_i[col])
            strictly_better |= (df[col] > row_i[col])
            
        for col in obj_min_cols:
            better_or_equal &= (df[col] <= row_i[col])
            strictly_better |= (df[col] < row_i[col])
            
        # If any other point is better or equal in all, AND strictly better in at least one
        if np.any(better_or_equal & strictly_better):
            pareto_mask[i] = False
            
    return df[pareto_mask]

# pareto_df = get_pareto_frontier(results_df, ['F1'], ['delta'])

Keep looking

Skills are one crate of 328,083. Ordering is by how many stacks a row turns up in, so the top of any crate is what has actually been picked rather than what has the most stars.