Papers

Filtered to on-policy self-distillation · clear filter

Browse by term

continual learning 83reinforcement learning 50large language models 12benchmarking 11benchmarks 11language models 11vision-language models 11robotics 7natural language processing 6world models 6generative models 5recursive self-improvement 5attention mechanisms 4diffusion Transformers 4multi-agent systems 4multimodal learning 4multimodal models 4on-policy distillation 4self-distillation 4self-supervised learning 4transformers 4video generation 4vision-language-action models 4agent-based systems 3agentic models 3agentic search 3autonomous systems 3coding agents 3diffusion models 3image generation 3

Matching papers

TTPO: Test-Time Policy Optimization

72 upvotes · 27 AUG 2026 · Aozhe Wang, Zhengxi Lu, Jianze Wang et al.

This paper proposes a new method called Test-Time Policy Optimization (TTPO) that allows large language models to be trained without ground-truth labels, enabling test-time training. Practitioners may care about this because it enables models to be trained without labels, which can be difficult or expensive to obtain.

reinforcement learningcontinual learningon-policy self-distillationtest-time trainingpseudo-labels

On-Policy Self-Distillation in Diffusion Models

47 upvotes · 25 AUG 2026 · Wei Zhou, Xiongwei Zhu, Lingdong Kong et al.

This paper introduces a new method for improving diffusion models by using self-distillation to align them with human preferences and task-specific objectives. Practitioners might care about this approach because it can lead to more efficient and analyzable alignment of diffusion models with human goals.

on-policy self-distillationdiffusion modelscontinual learningimage-level reward guidanceintermediate supervision