> ## Documentation Index
> Fetch the complete documentation index at: https://mintlify.com/skydiscover-ai/skydiscover/llms.txt
> Use this file to discover all available pages before exploring further.

# Search Configuration

> Configure evolutionary search algorithms and database settings in SkyDiscover

## Overview

The `search` section controls the evolutionary search algorithm, database backend, and context program selection strategy.

## Basic Configuration

```yaml theme={null}
search:
  type: "topk"              # Search algorithm type
  num_context_programs: 4   # Number of examples in prompt
  output_dir: null          # Override output directory
  database:                 # Algorithm-specific settings
    db_path: null
    log_prompts: true
```

## Search Types

SkyDiscover supports multiple search algorithms, each with different strategies for exploring the solution space.

### Top-K Search

Simple best-first search that maintains the top-k programs by fitness.

```yaml theme={null}
search:
  type: "topk"
  num_context_programs: 4
  database:
    db_path: "outputs/database.db"
    log_prompts: true
```

<ParamField path="db_path" type="str" default="None">
  Path to SQLite database for storing results
</ParamField>

<ParamField path="log_prompts" type="bool" default="true">
  Whether to log prompts to the database
</ParamField>

### AdaEvolve (Adaptive Multi-Island)

Adaptive evolutionary algorithm with multiple islands that adjust search intensity based on improvement history.

```yaml configs/adaevolve.yaml theme={null}
search:
  type: "adaevolve"
  num_context_programs: 4
  database:
    # Population settings
    population_size: 20
    num_islands: 2
    
    # Adaptive search intensity
    decay: 0.9
    intensity_min: 0.15
    intensity_max: 0.5
    
    # Feature flags
    use_adaptive_search: true
    use_ucb_selection: true
    use_migration: true
    use_unified_archive: true
    
    # Migration
    migration_interval: 15
    migration_count: 5
    
    # Archive settings
    fitness_weight: 1.0
    novelty_weight: 0.0
    diversity_strategy: "code"  # code, metric, or hybrid
    
    # Dynamic islands
    use_dynamic_islands: true
    max_islands: 5
    spawn_productivity_threshold: 0.015
    spawn_cooldown_iterations: 30
    
    # Paradigm breakthrough
    use_paradigm_breakthrough: true
    paradigm_window_size: 10
    paradigm_improvement_threshold: 0.12
    paradigm_max_uses: 2
    paradigm_max_tried: 10
    paradigm_num_to_generate: 3
```

Defined in `skydiscover/config.py:380-432`

<AccordionGroup>
  <Accordion title="Population Settings">
    <ParamField path="population_size" type="int" default="20">
      Number of programs per island
    </ParamField>

    <ParamField path="num_islands" type="int" default="2">
      Initial number of islands
    </ParamField>

    <ParamField path="archive_size" type="int" default="100">
      Maximum size of the unified archive
    </ParamField>
  </Accordion>

  <Accordion title="Adaptive Search">
    <ParamField path="use_adaptive_search" type="bool" default="true">
      Enable adaptive search intensity based on improvement signal
    </ParamField>

    <ParamField path="decay" type="float" default="0.9">
      Exponential moving average weight (rho) for improvement signal
    </ParamField>

    <ParamField path="intensity_min" type="float" default="0.15">
      Minimum search intensity (exploitation mode)
    </ParamField>

    <ParamField path="intensity_max" type="float" default="0.5">
      Maximum search intensity (exploration mode)
    </ParamField>

    <ParamField path="fixed_intensity" type="float" default="0.4">
      Fixed intensity when `use_adaptive_search=false`
    </ParamField>
  </Accordion>

  <Accordion title="Selection & Migration">
    <ParamField path="use_ucb_selection" type="bool" default="true">
      Use UCB (Upper Confidence Bound) for island selection
    </ParamField>

    <ParamField path="use_migration" type="bool" default="true">
      Enable inter-island migration
    </ParamField>

    <ParamField path="migration_interval" type="int" default="15">
      Migrate every N iterations
    </ParamField>

    <ParamField path="migration_count" type="int" default="5">
      Number of top programs to migrate
    </ParamField>
  </Accordion>

  <Accordion title="Archive & Diversity">
    <ParamField path="use_unified_archive" type="bool" default="true">
      Use quality-diversity archive
    </ParamField>

    <ParamField path="fitness_weight" type="float" default="1.0">
      Weight for fitness rank in elite score
    </ParamField>

    <ParamField path="novelty_weight" type="float" default="0.0">
      Weight for novelty rank in elite score
    </ParamField>

    <ParamField path="diversity_strategy" type="str" default="code">
      Diversity metric: `code`, `metric`, or `hybrid`
    </ParamField>

    <ParamField path="k_neighbors" type="int" default="5">
      Number of neighbors for novelty calculation
    </ParamField>
  </Accordion>

  <Accordion title="Dynamic Islands">
    <ParamField path="use_dynamic_islands" type="bool" default="true">
      Enable dynamic island spawning
    </ParamField>

    <ParamField path="max_islands" type="int" default="5">
      Maximum number of islands
    </ParamField>

    <ParamField path="spawn_productivity_threshold" type="float" default="0.015">
      Spawn new island if productivity drops below this
    </ParamField>

    <ParamField path="spawn_cooldown_iterations" type="int" default="30">
      Wait N iterations between island spawns
    </ParamField>
  </Accordion>

  <Accordion title="Paradigm Breakthrough">
    <ParamField path="use_paradigm_breakthrough" type="bool" default="true">
      Generate new high-level strategies during stagnation
    </ParamField>

    <ParamField path="paradigm_window_size" type="int" default="10">
      Window for improvement rate calculation
    </ParamField>

    <ParamField path="paradigm_improvement_threshold" type="float" default="0.12">
      Trigger paradigm breakthrough below this improvement rate
    </ParamField>

    <ParamField path="paradigm_max_uses" type="int" default="2">
      Maximum uses per paradigm
    </ParamField>

    <ParamField path="paradigm_num_to_generate" type="int" default="3">
      Number of paradigms to generate per trigger
    </ParamField>

    <ParamField path="paradigm_max_tried" type="int" default="10">
      Maximum tried paradigms to track
    </ParamField>
  </Accordion>
</AccordionGroup>

### OpenEvolve Native (MAP-Elites)

Quality-diversity search using MAP-Elites grids with island-based populations.

```yaml configs/openevolve_native.yaml theme={null}
search:
  type: "openevolve_native"
  num_context_programs: 5
  database:
    num_islands: 5
    population_size: 40
    archive_size: 100
    
    # Selection strategies
    exploration_ratio: 0.2    # P(explore) - random from island
    exploitation_ratio: 0.7   # P(exploit) - archive elite
    # remaining 0.1 = P(random)
    
    elite_selection_ratio: 0.1
    
    # MAP-Elites features
    feature_dimensions: ["complexity", "diversity"]
    feature_bins: 10
    diversity_reference_size: 20
    
    # Migration
    migration_interval: 10
    migration_rate: 0.1
    
    random_seed: 42
```

Defined in `skydiscover/config.py:435-450`

<ParamField path="num_islands" type="int" default="5">
  Number of islands in the population
</ParamField>

<ParamField path="population_size" type="int" default="40">
  Total population size across all islands
</ParamField>

<ParamField path="archive_size" type="int" default="100">
  Size of the MAP-Elites archive
</ParamField>

<ParamField path="exploration_ratio" type="float" default="0.2">
  Probability of exploration (random parent from current island)
</ParamField>

<ParamField path="exploitation_ratio" type="float" default="0.7">
  Probability of exploitation (elite from archive)
</ParamField>

<ParamField path="elite_selection_ratio" type="float" default="0.1">
  Fraction of context programs from top elites
</ParamField>

<ParamField path="feature_dimensions" type="List[str]" default="[&#x22;complexity&#x22;, &#x22;diversity&#x22;]">
  Behavioral dimensions for MAP-Elites grid
</ParamField>

<ParamField path="feature_bins" type="int" default="10">
  Number of bins per feature dimension
</ParamField>

### GEPA Native

Guided Evolution for Program Adaptation with elite pool, epsilon-greedy selection, and LLM-mediated merge.

```yaml theme={null}
search:
  type: "gepa_native"
  num_context_programs: 4
  database:
    population_size: 40
    
    # Selection
    candidate_selection_strategy: "epsilon_greedy"  # epsilon_greedy, best, pareto
    epsilon: 0.1
    max_rejection_history: 20
    
    # Controller settings
    acceptance_gating: true
    use_merge: true
    merge_after_stagnation: 15
    max_merge_attempts: 10
    max_recent_failures: 5
    
    random_seed: 42
```

Defined in `skydiscover/config.py:453-472`

### Beam Search

Beam search with diversity weighting.

```yaml theme={null}
search:
  type: "beam_search"
  num_context_programs: 4
  database:
    beam_width: 5
    beam_selection_strategy: "diversity_weighted"
    beam_diversity_weight: 0.3
    beam_temperature: 1.0
    beam_depth_penalty: 0.0
```

Defined in `skydiscover/config.py:362-370`

### Best-of-N

Generate N candidates and select the best one.

```yaml theme={null}
search:
  type: "best_of_n"
  num_context_programs: 4
  database:
    best_of_n: 5
```

Defined in `skydiscover/config.py:373-377`

### EvoX (Co-Evolution)

Label-guided co-evolutionary search.

```yaml configs/evox.yaml theme={null}
search:
  type: "evox"
  database:
    database_file_path: null  # Auto-configured
    evaluation_file: null     # Auto-configured
    config_path: null         # Auto-configured
    auto_generate_variation_operators: true
```

Defined in `skydiscover/config.py:339-359`

## SearchConfig Parameters

Defined in `skydiscover/config.py:485-493`

<ParamField path="type" type="str" default="topk">
  Search algorithm: `topk`, `adaevolve`, `openevolve_native`, `gepa_native`, `beam_search`, `best_of_n`, or `evox`
</ParamField>

<ParamField path="num_context_programs" type="int" default="4">
  Number of example programs to include in generation prompts
</ParamField>

<ParamField path="output_dir" type="str" default="None">
  Override output directory. If None, auto-generates based on search type and timestamp
</ParamField>

<ParamField path="database" type="DatabaseConfig" default="DatabaseConfig()">
  Algorithm-specific database configuration
</ParamField>

## CLI Overrides

Override search type from command line:

```bash theme={null}
# Use AdaEvolve
skydiscover-run program.py evaluator.py -s adaevolve

# Use OpenEvolve Native
skydiscover-run program.py evaluator.py -s openevolve_native

# Use beam search
skydiscover-run program.py evaluator.py -s beam_search
```

## Choosing a Search Algorithm

<CardGroup cols={2}>
  <Card title="Top-K" icon="list-ol">
    **Best for:** Quick experiments, simple optimization

    **Pros:** Fast, simple, low overhead

    **Cons:** Limited exploration
  </Card>

  <Card title="AdaEvolve" icon="island">
    **Best for:** Complex optimization, long runs

    **Pros:** Adaptive, robust, handles stagnation

    **Cons:** More computational overhead
  </Card>

  <Card title="OpenEvolve Native" icon="grid">
    **Best for:** Quality-diversity, exploring trade-offs

    **Pros:** Diverse solutions, explores full space

    **Cons:** Requires good feature dimensions
  </Card>

  <Card title="GEPA" icon="merge">
    **Best for:** Merging diverse solutions

    **Pros:** LLM-mediated combination, smart gating

    **Cons:** Higher LLM token usage
  </Card>
</CardGroup>

## Next Steps

<CardGroup cols={2}>
  <Card title="LLM Configuration" icon="brain" href="/config/llm">
    Configure models for search
  </Card>

  <Card title="Evaluator Configuration" icon="check-circle" href="/config/overview">
    Set up program evaluation
  </Card>
</CardGroup>