AI Paper Atlas

Research map / Reinforcement learning / Policy gradient and actor-critic

Separating value functions across time-scales

Joshua Romoff, Peter Henderson, Ahmed Touati, Emma Brunskill, Joelle Pineau, Yann Ollivier

ICML 2019 · Reinforcement learning › Policy gradient and actor-critic · approach: Temporal-difference and discounting theory · 2 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
Decompose PPO value estimation into discount-difference componentsPPO · Qbertreturn: 11735 -> 13428Improved
Decompose PPO value estimation into discount-difference componentsPPO · MsPacmanreturn: 1888 -> 2273Improved
Decompose PPO value estimation into discount-difference componentsPPO · Heroreturn: 21038 -> 29074Improved
Decompose PPO value estimation into discount-difference componentsPPO · Amidarreturn: 575 -> 731Improved
Decompose PPO value estimation into discount-difference componentsPPO · Zaxxonreturn: 7366 -> 396Hurt
Decompose PPO value estimation into discount-difference componentsPPO · WizardOfWorreturn: 3408 -> 2118Hurt
Limit each component's λ_z to 1 (ablation)PPO-TD(λ, Δ) · Zaxxonreturn: 396 -> 3291Improved
Limit each component's λ_z to 1 (ablation)PPO-TD(λ, Δ) · WizardOfWorreturn: 2118 -> 2440Improved
Use the fixed TD(λ, Δ) configuration rather than limiting λ_z to 1 (ablation)PPO-TD(λ̂, Δ) · Frostbitereturn: 304 -> 292No clear effect
Use the fixed TD(λ, Δ) configuration rather than limiting λ_z to 1 (ablation)PPO-TD(λ̂, Δ) · Qbertreturn: 13092 -> 13428No clear effect
Use an augmented network with summed outputs instead of TD(Δ) component targets (ablation)PPO · Heroreturn: 23511 -> 29074Hurt
Use shorter component-specific multi-step horizons (ablation)TD(Δ) · 5-state ring MDPabsolute errorImproved

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy