AsyncOPD: How Stale Can On-Policy Distillation Be?

• URL: http://arxiv.org/abs/2606.24143v1
• Date: Tue, 23 Jun 2026 04:50:49 GMT
• Title: AsyncOPD: How Stale Can On-Policy Distillation Be?
• Abstract summary: On-policy distillation (OPD) trains a student on its own rollouts guided by teacher feedback.<n>Asynchronous training pipelines can alleviate this bottleneck by decoupling rollout generation from learner updates.<n>We present the first systematic study of staleness in asynchronous OPD.
• Score: 23.273855195545394
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: On-policy distillation (OPD) trains a student on its own rollouts guided by teacher feedback and is becoming increasingly important for large language model (LLM) post-training. Like reinforcement learning (RL), however, OPD faces an on-policy systems bottleneck, as rollouts can dominate training time for reasoning workloads. Asynchronous training pipelines can alleviate this bottleneck by decoupling rollout generation from learner updates, but doing so introduces stale-policy data. While prior work has studied stale data in asynchronous RL, its effects in OPD remain underexplored. We present the first systematic study of staleness in asynchronous OPD, focusing on a practical setting where teacher feedback is implemented through local KL losses and full-vocabulary teacher logits are too expensive to store or transfer, necessitating finite teacher-score caches. We first show that KL direction changes the stale-data problem: teacher-weighted forward KL is more robust to stale rollouts, whereas student-weighted reverse KL is vulnerable. Second, for this vulnerable reverse-KL case, we study whether methods designed to stabilize asynchronous RL can mitigate OPD staleness. In our experiments, they do not improve over a simpler OPD-specific surrogate: recomputing the reverse-KL signal under the current student at learner time. Third, we analyze how finite teacher-score caches create a bias-variance tradeoff for sparse and sampled reverse-KL OPD estimators. This motivates multi-sample Monte Carlo (MC), which preserves MC correctability while reducing one-sample variance. Finally, we present and open-source AsyncOPD, a fully asynchronous OPD training pipeline built from these estimator choices. Experiments show that AsyncOPD improves training throughput by $1.6\times$ to $3.8\times$ over strict synchronous training while reaching comparable accuracy.

Related papers

• RL Forgets! Towards Continual Policy Optimization [57.4336338996653]
  Continual post-training is becoming a central paradigm for adapting vision-language models to evolving tasks.<n>Recent work has increasingly favored reinforcement learning over supervised fine-tuning, driven by the belief that reinforcement learning is inherently less prone to forgetting.<n>We propose Continual Policy Optimization (CPO), a replay-free framework grounded in a prior-task behavioral KL objective.
  arXiv  Detail & Related papers  (2026-07-05T15:46:11Z)
• Less is More: Early Stopping Rollout for On-Policy Distillation [39.392596318514244]
  Early Stopping Rollout (ESR) is a simple yet effective distillation strategy that simply restricts the rollout generation to the first response tokens.<n>We show that ESR both surpasses the full rollout OPD performance across model size, family, tasks and training regime.
  arXiv  Detail & Related papers  (2026-05-26T13:49:37Z)
• On-Policy Distillation with Best-of-N Teacher Rollout Selection [54.91780727674628]
  We propose BRTS, a Best-of-N Rollout Teacher Selection framework for on-policy distillation.<n>BRTS augments standard student-context OPD with a teacher-context supervision branch constructed from the teacher trajectory.<n>BRTS improves over standard OPD on challenging reasoning benchmarks, with the largest gains on harder datasets.
  arXiv  Detail & Related papers  (2026-05-10T19:49:00Z)
• Prune-OPD: Efficient and Reliable On-Policy Distillation for Long-Horizon Reasoning [66.52232008796294]
  Prune-OPD dynamically aligns training budgets with supervision quality.<n>It reduces training time by 37.6%--68.0% while preserving, and often improving, performance on challenging benchmarks.
  arXiv  Detail & Related papers  (2026-05-08T14:38:53Z)
• TCOD: Exploring Temporal Curriculum in On-Policy Distillation for Multi-turn Autonomous Agents [55.27396165691312]
  We identify a key limitation of vanilla OPD in multi-turn agent settings, which we term Trajectory-Level KL Instability.<n>We propose TCOD, a framework that controls the trajectory depth exposed to the student and progressively expands it with a curriculum schedule.<n> Experimental results across four student-teacher pairs show that TCOD mitigates KL escalation and enhances KL stability throughout training, improving agent performance by up to 18 points over vanilla OPD.
  arXiv  Detail & Related papers  (2026-04-27T03:38:27Z)
• Lightning OPD: Efficient Post-Training for Large Reasoning Models with Offline On-Policy Distillation [7.2992280064983825]
  On-policy distillation (OPD) has emerged as an efficient post-training paradigm for large language models.<n>Standard OPD requires a live teacher inference server throughout training, resulting in substantial infrastructure overhead.<n>We propose Lightning OPD, an offline on-policy distillation framework that enforces teacher consistency by precomputing teacher log-probabilities over SFT rollouts.
  arXiv  Detail & Related papers  (2026-04-14T17:44:50Z)
• Self-Distilled RLVR [57.37526213765131]
  We show that learning signals solely derived from the privileged teacher result in severe information leakage and unstable long-term training.<n>We propose textbfRLSD (textbfRLVR with textbfSelf-textbfDistillation)<n>This enables RLSD to simultaneously harness the strengths of both RLVR and OPSD, achieving a higher convergence ceiling and superior training stability.
  arXiv  Detail & Related papers  (2026-04-03T15:50:07Z)
• Revisiting On-Policy Distillation: Empirical Failure Modes and Simple Fixes [31.95045602299568]
  On-policy distillation (OPD) is appealing for large language model (LLM) post-training because it evaluates teacher feedback on student-generated rollouts rather than fixed teacher traces.<n>We revisit OPD from the estimator and implementation sides.<n>We identify three failure modes of sampled-token OPD: an imbalanced one-token signal, unreliable teacher guidance on student-generated prefixes, and distortions caused by tokenizer or special-token mismatch.
  arXiv  Detail & Related papers  (2026-03-26T15:35:59Z)
• Reinforcement-aware Knowledge Distillation for LLM Reasoning [63.53679456364683]
  Reinforcement learning (RL) post-training has recently driven gains in long chain-of-thought reasoning large language models (LLMs)<n>Most existing knowledge distillation methods are designed for supervised fine-tuning (SFT), relying on fixed teacher traces or teacher-student Kullback-Leibler (KL) divergence-based regularization.<n>We propose RL-aware distillation (RLAD), which performs selective imitation during RL -- guiding the student toward the teacher only when it improves the current policy update.
  arXiv  Detail & Related papers  (2026-02-26T00:20:39Z)
• CoPRIS: Efficient and Stable Reinforcement Learning via Concurrency-Controlled Partial Rollout with Importance Sampling [11.252930904797]
  We propose Concurrency- Controlled Partial Rollout with Importance Sampling (CoPRIS)<n>CoPRIS mitigates long-tail inefficiencies by maintaining a fixed number of concurrent rollouts, early-terminating once sufficient samples are collected, and reusing unfinished trajectories in subsequent rollouts.<n>Experiments show that CoPRIS achieves up to 1.94x faster training while maintaining comparable or superior performance to synchronous RL systems.
  arXiv  Detail & Related papers  (2025-11-05T11:39:32Z)

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.