# Retroactive Advantage Correction: Closed-Form V-Trace Bias Correction for Delay-Aware RLHF

arXiv: https://arxiv.org/abs/2606.27580 · HTML: https://mlsanity.com/paper/2606.27580/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Placed in Reward modeling and RLHF, one of 10 directions in LLM alignment / RLHF / preference optimisation.
- Builds on 24 of its 29 references that are in the map.
- First posted on arXiv 25 Jun 2026.

## Placed in

- LLM alignment / RLHF / preference optimisation / [Reward modeling and RLHF](https://mlsanity.com/alignment/reward-modeling-and-rlhf/)

## Builds on (most cited first, 24 of 24)

- [Training language models to follow instructions with human feedback](https://mlsanity.com/paper/2203.02155/) (2022)
- [Proximal Policy Optimization Algorithms](https://mlsanity.com/paper/1707.06347/) (2017)
- [DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Models](https://mlsanity.com/paper/2402.03300/) (2024)
- [Deep reinforcement learning from human preferences](https://mlsanity.com/paper/1706.03741/) (2017)
- [Learning to summarize from human feedback](https://mlsanity.com/paper/2009.01325/) (2020)
- [HybridFlow: A Flexible and Efficient RLHF Framework](https://mlsanity.com/paper/2409.19256/) (2024)
- [High-Dimensional Continuous Control Using Generalized Advantage Estimation](https://mlsanity.com/paper/1506.02438/) (2015)
- [LiveCodeBench: Holistic and Contamination Free Evaluation of Large Language Models for Code](https://mlsanity.com/paper/2403.07974/) (2024)
- [IMPALA: Scalable Distributed Deep-RL with Importance Weighted Actor-Learner Architectures](https://mlsanity.com/paper/1802.01561/) (2018)
- [Safe and Efficient Off-Policy Reinforcement Learning](https://mlsanity.com/paper/1606.02647/) (2016)
- [MiniMax-M1: Scaling Test-Time Compute Efficiently with Lightning Attention](https://mlsanity.com/paper/2506.13585/) (2025)
- [AReaL: A Large-Scale Asynchronous Reinforcement Learning System for Language Reasoning](https://mlsanity.com/paper/2505.24298/) (2025)
- [Asynchronous RLHF: Faster and More Efficient Off-Policy RL for Language Models](https://mlsanity.com/paper/2410.18252/) (2024)
- [Reinforcement Learning with Random Delays](https://mlsanity.com/paper/2010.02966/) (2020)
- [Prosperity before Collapse: How Far Can Off-Policy RL Reach with Stale Data on LLMs?](https://mlsanity.com/paper/2510.01161/) (2025)
- [BAPO: Stabilizing Off-Policy Reinforcement Learning for LLMs via Balanced Policy Optimization with Adaptive Clipping](https://mlsanity.com/paper/2510.18927/) (2025)
- [Laminar: A Scalable Asynchronous RL Post-Training Framework](https://mlsanity.com/paper/2510.12633/) (2025)
- [Off-Policy Reinforcement Learning with Delayed Rewards](https://mlsanity.com/paper/2106.11854/) (2021)
- [A-3PO: Accelerating Asynchronous LLM Training with Staleness-aware Proximal Policy Approximation](https://mlsanity.com/paper/2512.06547/) (2025)
- [Stable Asynchrony: Variance-Controlled Off-Policy RL for LLMs](https://mlsanity.com/paper/2602.17616/) (2026)
- [Multi-Agent Reinforcement Learning with Reward Delays](https://mlsanity.com/paper/2212.01441/) (2022)
- [Stabilizing Off-Policy Training for Long-Horizon LLM Agent via Turn-Level Importance Sampling and Clipping-Triggered Normalization](https://mlsanity.com/paper/2511.20718/) (2025)
- [GrandCode: Achieving Grandmaster Level in Competitive Programming via Agentic Reinforcement Learning](https://mlsanity.com/paper/2604.02721/) (2026)
- [OPPO: Accelerating PPO-based RLHF via Pipeline Overlap](https://mlsanity.com/paper/2509.25762/) (2025)

## Abstract (arXiv)

Reinforcement learning from human feedback (RLHF) in production does not always have a synchronous reward signal. Code-execution verifiers, slow judge ensembles, and queued human review can return several gradient steps after the rollout that produced them, breaking the synchronous-reward assumption underlying standard PPO. We address this gap with Retroactive Advantage Correction (RAC): each pending slow completion is queued, aged through a non-negative kernel, and reinjected as a clipped residual into the next optimiser step's advantage. We prove that under an unbiased clipped importance ratio, the cumulative RAC correction is exactly unbiased when the effective delay kernel reinjects all of its mass, and carries a bias linear in the unreinjected fraction otherwise; at the no-delay identity kernel it reduces to V-trace. On a tabular Markov decision process (MDP) proof-of-concept, RAC reduces the closed-form policy bias by up to 47.9x at the two-slow-channel configuration, beating wait-for-slow at lower wall-clock cost. RAC integrates with PPO and GRPO through a two-line reward-manager patch.