Struct AgentConfig 

pub struct AgentConfig {

    pub algorithm: String,
    pub ct_depth: usize,
    pub agent_horizon: usize,
    pub observation_bits: usize,
    pub observation_stream_len: usize,
    pub observation_key_mode: ObservationKeyMode,
    pub reward_bits: usize,
    pub agent_actions: usize,
    pub num_simulations: usize,
    pub exploration_exploitation_ratio: f64,
    pub discount_gamma: f64,
    pub min_reward: Reward,
    pub max_reward: Reward,
    pub reward_offset: Reward,
    pub rwkv_model_path: Option<String>,
    pub rosa_max_order: Option<i64>,
    pub zpaq_method: Option<String>,
}

Configuration parameters for an AIXI agent.

Fields

algorithm: String

The predictive algorithm to use (“ctw”, “rosa”, “rwkv”, “zpaq”).

ct_depth: usize

Context depth for the CTW model.

agent_horizon: usize

Planning horizon for MCTS.

observation_bits: usize

Number of bits used to encode observations.

observation_stream_len: usize

Number of observation symbols per action (stream length).

observation_key_mode: ObservationKeyMode

Strategy for mapping observation streams into search keys.

reward_bits: usize

Number of bits used to encode rewards.

agent_actions: usize

Number of possible actions.

num_simulations: usize

Number of MCTS simulations per planning step.

exploration_exploitation_ratio: f64

Constant governing exploration vs exploitation in UCT.

discount_gamma: f64

Discount factor for future rewards (1.0 = undiscounted).

min_reward: Reward

Minimum possible instantaneous reward in the environment.

max_reward: Reward

Maximum possible instantaneous reward in the environment.

reward_offset: Reward

Reward offset applied before encoding rewards as unsigned bits.

Paper-compatible encoding shifts rewards by an offset so all encoded values are non-negative.

rwkv_model_path: Option<String>

Path to the RWKV model weights (if using “rwkv”).

rosa_max_order: Option<i64>

Maximum Markov order for the ROSA model (if using “rosa”).

zpaq_method: Option<String>

ZPAQ method string for the rate model (if using “zpaq”).

Trait Implementations

impl Clone for AgentConfig

fn clone(&self) -> AgentConfig

Returns a duplicate of the value.

fn clone_from(&mut self, source: &Self)

Performs copy-assignment from source.

impl Debug for AgentConfig

fn fmt(&self, f: &mut Formatter<'_>) -> Result

Formats the value using the given formatter.