Offline RLAIF: Piloting VLM Feedback for RL via SFO

• URL: http://arxiv.org/abs/2503.01062v6
• Date: Sun, 14 Sep 2025 19:13:37 GMT
• Title: Offline RLAIF: Piloting VLM Feedback for RL via SFO
• Authors: Jacob Beck,
• Abstract summary: Vision-Language Models (VLMs) are limited in their ability to solve control tasks due to their lack of action-conditioned training data.<n>A key challenge in Reinforcement Learning from AI Feedback is determining how best to integrate VLM-derived signals into the learning process.
• Score: 4.391505380846452
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: While internet-scale image and textual data have enabled strong generalization in Vision-Language Models (VLMs), the absence of internet-scale control data has impeded the development of similar generalization in standard reinforcement learning (RL) agents. Although VLMs are fundamentally limited in their ability to solve control tasks due to their lack of action-conditioned training data, their capacity for image understanding allows them to provide valuable feedback in RL tasks by recognizing successful outcomes. A key challenge in Reinforcement Learning from AI Feedback (RLAIF) is determining how best to integrate VLM-derived signals into the learning process. We explore this question in the context of offline RL and introduce a class of methods called Sub-Trajectory Filtered Optimization (SFO). We identify three key insights. First, trajectory length plays a crucial role in offline RL, as full-trajectory preference learning exacerbates the stitching problem, necessitating the use of sub-trajectories. Second, even in Markovian environments, a non-Markovian reward signal from a sequence of images is required to assess trajectory improvement, as VLMs do not interpret control actions and must rely on visual cues over time. Third, a simple yet effective approach--filtered and weighted behavior cloning--consistently outperforms more complex RLHF-based methods. We propose Sub-Trajectory Filtered Behavior Cloning (SFBC), a method that leverages VLM feedback on sub-trajectories while incorporating a retrospective filtering mechanism that removes sub-trajectories preceding failures to improve robustness and prevent turbulence. Please enjoy our airport puns.

Related papers

• Found-RL: foundation model-enhanced reinforcement learning for autonomous driving [15.275134927543611]
  Reinforcement Learning (RL) has emerged as a dominant paradigm for end-to-end autonomous driving (AD)<n>Found-RL is a platform tailored to efficiently enhance RL for AD using foundation models.<n>A core innovation is the asynchronous batch inference framework, which decouples heavy VLM reasoning from the simulation loop.
  arXiv  Detail & Related papers  (2026-02-11T02:56:04Z)
• COVR:Collaborative Optimization of VLMs and RL Agent for Visual-Based Control [37.27710513408952]
  Visual reinforcement learning (RL) suffers from poor sample efficiency due to high-dimensional observations in complex tasks.<n>We propose COVR, a collaborative optimization framework that enables the mutual enhancement of the VLM and RL policies.
  arXiv  Detail & Related papers  (2026-01-04T03:53:05Z)
• Expressive Value Learning for Scalable Offline Reinforcement Learning [9.946269411850064]
  Reinforcement learning (RL) is a powerful paradigm for learning to make sequences of decisions.<n> offline RL offers a promising avenue by training agents on large, diverse datasets.<n>We introduce Expressive Value Learning for Offline Reinforcement Learning (EVOR), a scalable offline RL approach that integrates both expressive policies and expressive value functions.
  arXiv  Detail & Related papers  (2025-10-09T13:42:20Z)
• Beyond Accuracy: Dissecting Mathematical Reasoning for LLMs Under Reinforcement Learning [93.00629872970364]
  Reinforcement learning (RL) has become the dominant paradigm for improving the performance of language models on complex reasoning tasks.<n>We introduce SPARKLE, a fine-grained analytic framework to dissect the effects of RL across three key dimensions.<n>We study whether difficult problems -- those yielding no RL signals and mixed-quality reasoning traces -- can still be effectively used for training.
  arXiv  Detail & Related papers  (2025-06-05T07:53:59Z)
• ViaRL: Adaptive Temporal Grounding via Visual Iterated Amplification Reinforcement Learning [68.76048244253582]
  We introduce ViaRL, the first framework to leverage rule-based reinforcement learning (RL) for optimizing frame selection in video understanding.<n>ViaRL utilizes the answer accuracy of a downstream model as a reward signal to train a frame selector through trial-and-error.<n>ViaRL consistently delivers superior temporal grounding performance and robust generalization across diverse video understanding tasks.
  arXiv  Detail & Related papers  (2025-05-21T12:29:40Z)
• RAGEN: Understanding Self-Evolution in LLM Agents via Multi-Turn Reinforcement Learning [125.65034908728828]
  Training large language models (LLMs) as interactive agents presents unique challenges. While reinforcement learning has enabled progress in static tasks, multi-turn agent RL training remains underexplored. We propose StarPO, a general framework for trajectory-level agent RL, and introduce RAGEN, a modular system for training and evaluating LLM agents.
  arXiv  Detail & Related papers  (2025-04-24T17:57:08Z)
• Exploring the Effect of Reinforcement Learning on Video Understanding: Insights from SEED-Bench-R1 [53.894789613838654]
  We introduce SEED-Bench-R1, a benchmark designed to evaluate post-training methods for MLLMs in video understanding.<n>It includes intricate real-world videos and complex everyday planning tasks in the format of multiple-choice questions.<n>Using Qwen2-VL-Instruct-7B as a base model, we compare RL with supervised fine-tuning (SFT)<n>Our detailed analysis reveals that RL enhances visual perception but often produces less coherent reasoning chains.
  arXiv  Detail & Related papers  (2025-03-31T17:55:23Z)
• Online Preference-based Reinforcement Learning with Self-augmented Feedback from Large Language Model [17.4036850872656]
  Preference-based reinforcement learning (PbRL) provides a powerful paradigm to avoid meticulous reward engineering by learning rewards based on human preferences.<n>In this paper, we propose a RL Self-augmented Large Language Model Feedback (RL-SaLLM-F) technique that does not rely on privileged information for online PbRL.
  arXiv  Detail & Related papers  (2024-12-22T06:15:25Z)
• An Examination of Offline-Trained Encoders in Vision-Based Deep Reinforcement Learning for Autonomous Driving [0.0]
  Research investigates the challenges Deep Reinforcement Learning (DRL) faces in Partially Observable Markov Decision Processes (POMDP) Our research adopts an offline-trained encoder to leverage large video datasets through self-supervised learning to learn generalizable representations. We show that the features learned by watching BDD100K driving videos can be directly transferred to achieve lane following and collision avoidance in CARLA simulator.
  arXiv  Detail & Related papers  (2024-09-02T14:16:23Z)
• MarvelOVD: Marrying Object Recognition and Vision-Language Models for Robust Open-Vocabulary Object Detection [107.15164718585666]
  We investigate the root cause of VLMs' biased prediction under the open vocabulary detection context. Our observations lead to a simple yet effective paradigm, coded MarvelOVD, that generates significantly better training targets. Our method outperforms the other state-of-the-arts by significant margins.
  arXiv  Detail & Related papers  (2024-07-31T09:23:57Z)
• How Can LLM Guide RL? A Value-Based Approach [68.55316627400683]
  Reinforcement learning (RL) has become the de facto standard practice for sequential decision-making problems by improving future acting policies with feedback. Recent developments in large language models (LLMs) have showcased impressive capabilities in language understanding and generation, yet they fall short in exploration and self-improvement capabilities. We develop an algorithm named LINVIT that incorporates LLM guidance as a regularization factor in value-based RL, leading to significant reductions in the amount of data needed for learning.
  arXiv  Detail & Related papers  (2024-02-25T20:07:13Z)
• Reasoning with Latent Diffusion in Offline Reinforcement Learning [11.349356866928547]
  offline reinforcement learning holds promise as a means to learn high-reward policies from a static dataset. Key challenge in offline RL lies in effectively stitching portions of suboptimal trajectories from the static dataset. We propose a novel approach that leverages the expressiveness of latent diffusion to model in-support trajectory sequences as compressed latent skills.
  arXiv  Detail & Related papers  (2023-09-12T20:58:21Z)
• Contextual Transformer for Offline Meta Reinforcement Learning [16.587320914107128]
  We show how prompts can improve sequence modeling-based offline reinforcement learning ( offline RL) algorithms. We propose prompt tuning for offline RL, where a context vector sequence istextuald with the input to guide the conditional policy generation. We extend our framework to Meta-RL settings and propose Contextual Meta Transformer (CMT); CMT leverages the context among different tasks as the prompt to improve generalization on unseen tasks.
  arXiv  Detail & Related papers  (2022-11-15T10:00:14Z)
• Mastering the Unsupervised Reinforcement Learning Benchmark from Pixels [112.63440666617494]
  Reinforcement learning algorithms can succeed but require large amounts of interactions between the agent and the environment. We propose a new method to solve it, using unsupervised model-based RL, for pre-training the agent. We show robust performance on the Real-Word RL benchmark, hinting at resiliency to environment perturbations during adaptation.
  arXiv  Detail & Related papers  (2022-09-24T14:22:29Z)
• Challenges and Opportunities in Offline Reinforcement Learning from Visual Observations [58.758928936316785]
  offline reinforcement learning from visual observations with continuous action spaces remains under-explored. We show that modifications to two popular vision-based online reinforcement learning algorithms suffice to outperform existing offline RL methods.
  arXiv  Detail & Related papers  (2022-06-09T22:08:47Z)
• RvS: What is Essential for Offline RL via Supervised Learning? [77.91045677562802]
  Recent work has shown that supervised learning alone, without temporal difference (TD) learning, can be remarkably effective for offline RL. In every environment suite we consider simply maximizing likelihood with two-layer feedforward is competitive. They also probe the limits of existing RvS methods, which are comparatively weak on random data.
  arXiv  Detail & Related papers  (2021-12-20T18:55:16Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.