ConSFT prevents catastrophic forgetting in fine-tuning flow-matching VLAs by dynamically scaling gradients based on model confidence, retaining over 20% more pre-trained capability than standard SFT without prior data or reference networks.
arXiv preprint arXiv:2507.21053 , year=
9 Pith papers cite this work. Polarity classification is still indexing.
9
Pith papers citing it
years
2026 9verdicts
UNVERDICTED 9representative citing papers
SoftGAC defines a stochastic bridge from base to action latent that converts the MaxEnt objective into a tractable relative-entropy term reducible to control energy, achieving competitive returns with one-pass sampling.
DIAL uses intent-conditioned CFG and multi-intent GRPO to expand and preserve diverse modes in continuous-action preference RL, lifting RFS to 9.14 and surpassing both prior best (8.5) and human demonstration (8.13).
DRIFT enables stable offline-to-online fine-tuning of CTMC policies in discrete RL via advantage-weighted discrete flow matching, path-space regularization, and candidate-set approximation.
UniSteer unifies human corrective actions and noise-space RL for VLA adaptation by inverting actions to noise targets, raising success rates from 20% to 90% in 66 minutes across four real-world manipulation tasks.
OGPO is a sample-efficient off-policy method for full finetuning of generative control policies that reaches SOTA on robotic manipulation tasks and can recover from poor behavior-cloning initializations without expert data.
V-GRPO makes ELBO surrogates stable and efficient for online RL alignment of denoising models, delivering SOTA text-to-image performance with 2-3x speedups over MixGRPO and DiffusionNFT.
PODPO is a likelihood-free generative policy optimization method for online RL that steers actions to high-return regions using only positive-advantage samples and local contrastive drifting.
Sol-RL decouples FP4-based candidate exploration from BF16 policy optimization in diffusion RL, delivering up to 4.64x faster convergence with maintained or superior alignment performance on models like FLUX.1 and SD3.5.
citing papers explorer
-
Preserving Foundational Capabilities in Flow-Matching VLAs through Conservative SFT
ConSFT prevents catastrophic forgetting in fine-tuning flow-matching VLAs by dynamically scaling gradients based on model confidence, retaining over 20% more pre-trained capability than standard SFT without prior data or reference networks.
-
Generative Actor-Critic with Soft Bridge Policies
SoftGAC defines a stochastic bridge from base to action latent that converts the MaxEnt objective into a tractable relative-entropy term reducible to control energy, achieving competitive returns with one-pass sampling.
-
Driving Intents Amplify Planning-Oriented Reinforcement Learning
DIAL uses intent-conditioned CFG and multi-intent GRPO to expand and preserve diverse modes in continuous-action preference RL, lifting RFS to 9.14 and surpassing both prior best (8.5) and human demonstration (8.13).
-
Discrete Flow Matching for Offline-to-Online Reinforcement Learning
DRIFT enables stable offline-to-online fine-tuning of CTMC policies in discrete RL via advantage-weighted discrete flow matching, path-space regularization, and candidate-set approximation.
-
Unified Noise Steering for Efficient Human-Guided VLA Adaptation
UniSteer unifies human corrective actions and noise-space RL for VLA adaptation by inverting actions to noise targets, raising success rates from 20% to 90% in 66 minutes across four real-world manipulation tasks.
-
OGPO: Sample Efficient Full-Finetuning of Generative Control Policies
OGPO is a sample-efficient off-policy method for full finetuning of generative control policies that reaches SOTA on robotic manipulation tasks and can recover from poor behavior-cloning initializations without expert data.
-
V-GRPO: Online Reinforcement Learning for Denoising Generative Models Is Easier than You Think
V-GRPO makes ELBO surrogates stable and efficient for online RL alignment of denoising models, delivering SOTA text-to-image performance with 2-3x speedups over MixGRPO and DiffusionNFT.
-
Positive-Only Drifting Policy Optimization
PODPO is a likelihood-free generative policy optimization method for online RL that steers actions to high-return regions using only positive-advantage samples and local contrastive drifting.
-
FP4 Explore, BF16 Train: Diffusion Reinforcement Learning via Efficient Rollout Scaling
Sol-RL decouples FP4-based candidate exploration from BF16 policy optimization in diffusion RL, delivering up to 4.64x faster convergence with maintained or superior alignment performance on models like FLUX.1 and SD3.5.