gepa-spo/dist/bandit.js

Genetic-Pareto prompt optimizer to evolve system prompts from a few rollouts with modular support and intelligent crossover

/** UCB1 bandit to select strategy IDs by historical uplift */
export class UCB1 {
    t = 0;
    stats;
    constructor(ids) {
        this.stats = ids.map(id => ({ id, n: 0, mean: 0 }));
    }
    pick() {
        this.t++;
        const c = Math.sqrt(2);
        let bestId = this.stats[0]?.id ?? '';
        let bestU = -Infinity;
        for (const s of this.stats) {
            const bonus = s.n ? c * Math.sqrt(Math.log(this.t) / s.n) : Number.POSITIVE_INFINITY;
            const u = (s.n ? s.mean : 0) + bonus;
            if (u > bestU) {
                bestU = u;
                bestId = s.id;
            }
        }
        return bestId;
    }
    update(id, reward) {
        const s = this.stats.find(x => x.id === id);
        if (!s)
            return;
        const r = Math.max(0, Math.min(1, reward));
        s.n += 1;
        s.mean += (r - s.mean) / s.n;
    }
    serialize() { return { t: this.t, stats: this.stats.map(s => ({ ...s })) }; }
    static from(obj) { const b = new UCB1([]); b.t = obj.t; b.stats = obj.stats.map(s => ({ ...s })); return b; }
}