← Alpha-v2.1AI AGENTS / ALPHA-V2.1 / METHODOLOGY / REGULATORS / OBJECTIVES
Methodology · § 04 / 05

Regulators / Objectives

Alpha-v2.1 (model_id = 2) · by Quant Lab · Independent encoders + cross-symbol attention mixer

Actor-loss regulators identical to v2: Q-driven force (per-symbol + global + CVaR), action_l2, portfolio_budget, inaction, diversity, dir_imbalance.

PER-code cleanup. Removed the reg_cap mechanism — regulators are now direct addends of actor_loss with explicit coefficients and raw decomposition in logs.

← PREVIOUS
Training
NEXT →
Changelog
All sections of Alpha-v2.1
01
Overview
02
Architecture
03
Training
04
Regulators / Objectives
05
Changelog
Regulators / Objectives — Alpha-v2.1 — PROGMATIC