Are Full Rollouts Necessary for On-Policy Distillation?
This paper investigates whether full rollouts are required during on-policy distillation (OPD) for training reasoning models, identifying rollout horizon as a key computational bottleneck. The authors propose two strategies: Progressive OPD (POPD), which gradually expands rollout horizon during training, and Truncated OPD (TOPD), which uses permanently truncated rollouts. Experiments on mathematical reasoning show POPD achieves up to 3× training efficiency improvement, while TOPD matches full OPD performance using only 10% of the rollout horizon, yielding significant wall-clock and memory savings.
Related guides (3)
Related events (8)
Analysis of on-policy distillation reveals sparse, geometrically structured parameter updates
A new arXiv paper analyzes on-policy distillation (OPD) — a post-training method combining on-policy student trajectories with dense teacher supervision — across language and vision-language model pairs. The authors find that OPD updates are coordinate-sparse and distributed across layers (FFN-heavy), and that training only the discovered sparse subnetwork recovers near-full performance. Geometrically, updates are numerically full-rank but spectrally concentrated, falling disproportionately on near-zero weight coordinates, suggesting OPD retains distinct geometric signatures rather than behaving like ordinary dense parameter rewriting.
OPCoD: On-Policy Co-Distillation for Mutual LLM Improvement via Peer Feedback
Researchers introduce On-Policy Co-Distillation (OPCoD), a training framework where two LLMs, each stronger in a different domain, iteratively tutor each other using on-policy rollouts and peer feedback. The method uses cognizance-based gating to control when feedback is given and feedback anchoring to ground it in the problem context. On Science Q&A tasks, OPCoD achieves Pareto improvement for both models across all evaluated domain pairs, outperforming one-way distillation and single-model fine-tuning baselines.
d-OPSD: First on-policy self-distillation framework tailored for diffusion LLMs
Researchers introduce d-OPSD, the first on-policy self-distillation (OPSD) framework designed specifically for diffusion large language models (dLLMs). The method addresses a fundamental mismatch between existing autoregressive OPSD approaches and dLLMs' arbitrary-order generation by using suffix conditioning on self-generated answers and step-level rather than token-level divergence supervision. Across four reasoning benchmarks, d-OPSD outperforms RLVR and SFT baselines while requiring only ~10% of the optimization steps of RLVR, suggesting strong sample efficiency gains for dLLM post-training.
Canonical-Context On-Policy Distillation (CCOPD) for Multi-Turn LLM Consistency
This paper identifies 'self-anchored drift' as a key failure mode in multi-turn LLMs: when information is revealed incrementally across turns, models produce unsupported assumptions that distort final answers, even when the total evidence is identical to a single-prompt setting. The authors propose Canonical-Context On-Policy Distillation (CCOPD), which trains a student model on incremental multi-turn conversations to match the output distribution of a frozen teacher conditioned on the full clean prompt. Trained only on math conversations, CCOPD achieves a 32% average relative improvement on multi-turn (RAW-SHARDED) tasks and generalizes zero-shot to five out-of-domain task families while preserving single-prompt performance.
DRPO: Smooth divergence regularization replaces hard masking in LLM RL training
A new arXiv preprint proposes Divergence Regularized Policy Optimization (DRPO), a method that replaces the hard trust-region mask used in DPPO with a smooth advantage-weighted quadratic regularizer on policy shift. The approach addresses a known weakness in PPO and GRPO where importance ratios poorly proxy distributional shift in long-tailed vocabularies, and in DPPO where gradient signals are discarded rather than corrected at trust-region boundaries. Experiments across model scales, architectures, and precision settings show improved stability and efficiency in LLM RL post-training.
Vision-OPD: On-Policy Self-Distillation for Fine-Grained Visual Understanding in MLLMs
Vision-OPD addresses a 'regional-to-global perception gap' in multimodal LLMs, where models answer fine-grained visual questions more accurately when given cropped evidence regions than full images. The method instantiates a crop-conditioned teacher and full-image-conditioned student from the same MLLM, minimizing token-level divergence along on-policy rollouts to transfer regional perception to the full-image policy. This self-distillation requires no external teacher models, ground-truth labels, reward verifiers, or inference-time tools. Benchmarks show competitive or superior performance against larger open-source, closed-source, and agentic 'Thinking-with-Images' models.
GraphPO: Graph-based Policy Optimization reduces redundancy in LLM reasoning RL
GraphPO is a new reinforcement learning framework that represents reasoning rollouts as directed acyclic graphs rather than independent chains or trees, merging semantically equivalent reasoning paths into equivalence classes to share suffixes and reduce redundant exploration. The approach assigns efficiency advantages to incoming edges and correctness advantages to outgoing edges, deriving process supervision from outcome rewards. Experiments on three LLMs across reasoning and agentic search benchmarks show consistent improvements over chain- and tree-based baselines under equal token or response budgets. The method also provides theoretical guarantees on reduced advantage-estimation variance.
ZPPO: Teacher-in-prompt training method outperforms distillation and GRPO for small vision-language models
Researchers introduce Zone of Proximal Policy Optimization (ZPPO), a training method inspired by Vygotsky's zone of proximal development that embeds teacher guidance in prompts rather than policy gradients or logit imitation. On hard questions where student rollouts fail, ZPPO constructs Binary Candidate-included Questions (BCQ) and Negative Candidate-included Questions (NCQ) to help the student discriminate correct from incorrect responses, with a replay buffer that recirculates hard questions until mastered. Evaluated on the Qwen3 family (0.8B–9B) with a 27B teacher across a 31-benchmark suite covering VLM, LLM, and video tasks, ZPPO outperforms both distillation and GRPO baselines, with the largest gains at the smallest model scale. The method addresses a known failure mode of RL training where zero-reward rollouts produce no gradient signal.