Ethan Park

Designs self-consistency sampling strategies with optimal parameters for improved answer reliability.

Generates solutions using tree-of-thought exploration with configurable search strategies and backtracking.

Optimizes prompts to fit within token budgets while preserving priority elements and semantic meaning.

Creates self-improving prompt systems that analyze performance and generate optimized versions automatically.

Compares prompt versions to analyze changes, predict behavioral impact, and recommend A/B testing strategies.

Analyzes problems from multiple role-based perspectives to generate comprehensive, multi-viewpoint insights.

Validates chain-of-thought reasoning for logical validity, unsupported assumptions, and conclusion support.

Curates optimal few-shot example sets by analyzing coverage, difficulty distribution, and learning progression.

Identifies potential edge cases and failure modes in prompts without requiring example failures.

Designs optimal output formats for various content types with parsing specifications and validation markers.

Implements reflection loops to analyze prompt performance and generate improved versions based on failure analysis.

Breaks complex tasks into atomic subtasks with dependency mapping and parallel execution planning.