arXiv subjects

Zhenxuan Pan

Publications and source records attributed to Zhenxuan Pan.

7 recordsLinked to original sources

arXiv2026

AInfer-PD: Communication-Safe In-Place Prefill-Decode Multiplexing for Distributed MoE Rollouts

Rollout inference often dominates the wall-clock time of large-scale reinforcement learning (RL). In agentic RL, each trajectory alternates between model generation and environment interaction over multiple turns. Asynchronous trajectories consequently introduce new prefill (P) work while other trajectories remain in decode (D), making P/D coexistence a persistent property of the rollout rather than a one-time prompt-ingestion event. On shared accelerators, persistent P/D coexistence can make prefill interfere with latency-sensitive decode and prolong rollout completion. P/D disaggregation avoids this co-location but requires separate device pools and KV-cache transfers. In-place multiplexing retains shared devices and KV state, but existing designs lack the communication isolation needed for large MoE deployments that combine attention TP/DP with distributed expert execution. In practical implementations, P and D can issue intersecting collectives in inconsistent cross-rank orders; DeepEP's P and D paths also share mutable protocol state. We present AInfer-PD, which extends in-place P/D multiplexing to distributed MoE rollouts. AInfer-PD coordinates P/D collective order across ranks and gives the two DeepEP paths independent communication state, making crossed ADP/ATP and DeepEP paths safe for concurrent P/D execution. The design retains shared model weights and KV storage while coordinating P and D on the same devices. Across repeated single-node prefill-intensive workloads, AInfer-PD reduces fixed-workload rollout completion time by 7.1-22.5% relative to the same AInfer engine with P/D multiplexing disabled and by 24.8-32.9% relative to SGLang. On two nodes, the reductions are 18.0-35.3% and 18.3-31.8%, respectively. In a same-engine ablation, fine-grained boundaries reduce completion time by a further 8.6-19.8% over whole-epoch asynchronous enqueue.

cs.DC↗

arXiv2026

HARTS: Efficient Agentic Reinforcement Learning for Hybrid-Attention Models over Arbitrary Rollout Trees

Agentic reinforcement learning (RL) often produces irregular rollout trees with shared histories. Training root-to-leaf trajectories independently recomputes these shared prefixes. Existing systems primarily target full-attention models and lack dense, differentiable hybrid-attention execution compatible with activation recomputation. We present HARTS (Hybrid-Attention RL over Tree Structures). HARTS jointly plans microbatches, data-parallel (DP) replica assignments, and microbatch-slot schedules using non-replay compact-token work after prefix compression. For chunkwise linear attention, a linear-time algorithm coordinates chunk-boundary state recovery and replay and produces the minimum number of sequential linear-attention calls under our packed execution model. HARTS preserves the chunkwise state partitioning of trajectory-wise training: it does not repeat projections, MLP/MoE computation, or final outputs, and performs only bounded state replay for numerical alignment. Per round, HARTS batches all branches into one packed call, propagates gradients through differentiable state handoffs, supports activation recomputation, and restores per-token log-probabilities. For deterministic, no-token-drop top-$k$ MoE routing, semantic multiplicities restore MoE-objective token weights and load statistics. Existing RL objectives retain their interface. To our knowledge, HARTS is the first system to demonstrate arbitrary-rollout-tree prefix-sharing speedups on a real hybrid-attention model. On an Agentic RL workload generated from SWE-bench tasks, HARTS achieves $4.81$--$4.87\times$ forward/backward/gradient speedup with activation recomputation across multiple parallel configurations. Its numerical differences are comparable to baseline self-rerun variation, and its reward trend is similar to the baseline over the first 120 steps of $\tau^3$-Bench training.

cs.LG↗

arXiv2026

RoutePack: Expert Placement and Attention-Aware Data Packing for MoE Reinforcement Learning

Training Mixture-of-Experts (MoE) models for reinforcement learning (RL) couples two load-balancing problems: sequence composition determines dense attention work in each data-parallel microbatch, while token routing determines sparse expert work on expert-parallel ranks. Optimizing either alone can shift the bottleneck to the other. In MoE RL, rollout-time routing replay exposes every sample's sequence length and layer-wise expert demand before its training step. We present RoutePack, a hierarchical planner that coordinates state-consistent, layer-wise expert rerouting with joint attention- and expert-aware data packing over an optimizer-step window. RoutePack first places experts independently at each MoE layer using aggregate routing demand. It then packs samples into the smallest certified, or best-known feasible, number of token-capped execution rows and optimizes their DP layout with a projected EDP-shard-aware objective. The objective combines a window-normalized linear-quadratic attention proxy with per-layer physical EP-rank peaks and minimizes the accumulated cost of the slowest EDP shard. Parallel population annealing searches fixed-row feasible layouts while preserving sample coverage, capacity, nonempty cells, equal microbatch counts, and communicator topology. State-consistent materialization preserves logical top-k routing and existing MoE kernels without microbatch-level expert replication. Across Ling-3.0-Tiny and Ling-3.0-Flash, expert rerouting improves mean trainer-measured token throughput by 3.80% and 10.50%, while routing-aware packing adds another 4.86% and 3.98%, respectively. Overall, RoutePack improves throughput by 8.85% and 14.89% over the baseline.

cs.DC↗

arXiv2026

TAD: Temporal-Aware Trajectory Self-Distillation for Fast and Accurate Diffusion LLM

Diffusion large language models (dLLMs) offer a promising paradigm for parallel text generation, but in practice they face an accuracy-parallelism trade-off, where increasing tokens per forward (TPF) often degrades generation quality. Existing acceleration methods often gain speed at the cost of accuracy. To address this limitation, we propose TAD, a Temporal-Aware trajectory self-Distillation framework. During data construction, we condition a teacher model on both the prompt and the ground-truth response to generate decoding trajectories, recording the intermediate masked states throughout the process. Based on how many decoding steps remain before each masked token is revealed, we partition masked positions into near and distant subsets. For near tokens, we train the student with a hard cross-entropy loss using the teacher trajectory tokens as labels, encouraging confident predictions for tokens that are about to be decoded. For distant tokens, we apply a soft KL divergence loss between the teacher and student token distributions, providing softer supervision and preserving future planning knowledge. This temporal-aware partition naturally gives rise to two deployment configurations: a Quality model that prioritizes accuracy and a Speed model that favors more aggressive acceleration. Experiments show that TAD consistently improves the accuracy-parallelism trade-off. On LLaDA, it raises average accuracy from 46.2\% to 51.6\% with the Quality model and average AUP from 46.2 to 257.1 with the Speed model. Our code is available at: https://github.com/BHmingyang/TAD

cs.CL↗

arXiv2026

LLaDA2.1: Speeding Up Text Diffusion via Token Editing

While LLaDA2.0 showcased the scaling potential of 100B-level block-diffusion models and their inherent parallelization, the delicate equilibrium between decoding speed and generation quality has remained an elusive frontier. Today, we unveil LLaDA2.1, a paradigm shift designed to transcend this trade-off. By seamlessly weaving Token-to-Token (T2T) editing into the conventional Mask-to-Token (M2T) scheme, we introduce a joint, configurable threshold-decoding scheme. This structural innovation gives rise to two distinct personas: the Speedy Mode (S Mode), which audaciously lowers the M2T threshold to bypass traditional constraints while relying on T2T to refine the output; and the Quality Mode (Q Mode), which leans into conservative thresholds to secure superior benchmark performances with manageable efficiency degrade. Furthering this evolution, underpinned by an expansive context window, we implement the first large-scale Reinforcement Learning (RL) framework specifically tailored for dLLMs, anchored by specialized techniques for stable gradient estimation. This alignment not only sharpens reasoning precision but also elevates instruction-following fidelity, bridging the chasm between diffusion dynamics and complex human intent. We culminate this work by releasing LLaDA2.1-Mini (16B) and LLaDA2.1-Flash (100B). Across 33 rigorous benchmarks, LLaDA2.1 delivers strong task performance and lightning-fast decoding speed. Despite its 100B volume, on coding tasks it attains an astounding 892 TPS on HumanEval+, 801 TPS on BigCodeBench, and 663 TPS on LiveCodeBench.

cs.LG↗

arXiv2025

Every Step Evolves: Scaling Reinforcement Learning for Trillion-Scale Thinking Model

We present Ring-1T, the first open-source, state-of-the-art thinking model with a trillion-scale parameter. It features 1 trillion total parameters and activates approximately 50 billion per token. Training such models at a trillion-parameter scale introduces unprecedented challenges, including train-inference misalignment, inefficiencies in rollout processing, and bottlenecks in the RL system. To address these, we pioneer three interconnected innovations: (1) IcePop stabilizes RL training via token-level discrepancy masking and clipping, resolving instability from training-inference mismatches; (2) C3PO++ improves resource utilization for long rollouts under a token budget by dynamically partitioning them, thereby obtaining high time efficiency; and (3) ASystem, a high-performance RL framework designed to overcome the systemic bottlenecks that impede trillion-parameter model training. Ring-1T delivers breakthrough results across critical benchmarks: 93.4 on AIME-2025, 86.72 on HMMT-2025, 2088 on CodeForces, and 55.94 on ARC-AGI-1. Notably, it attains a silver medal-level result on the IMO-2025, underscoring its exceptional reasoning capabilities. By releasing the complete 1T parameter MoE model to the community, we provide the research community with direct access to cutting-edge reasoning capabilities. This contribution marks a significant milestone in democratizing large-scale reasoning intelligence and establishes a new baseline for open-source model performance.

cs.CL↗

arXiv2024

LayerKV: Optimizing Large Language Model Serving with Layer-wise KV Cache Management

The expanding context windows in large language models (LLMs) have greatly enhanced their capabilities in various applications, but they also introduce significant challenges in maintaining low latency, particularly in Time to First Token (TTFT). This paper identifies that the sharp rise in TTFT as context length increases is predominantly driven by queuing delays, which are caused by the growing demands for GPU Key-Value (KV) cache allocation clashing with the limited availability of KV cache blocks. To address this issue, we propose LayerKV, a simple yet effective plug-in method that effectively reduces TTFT without requiring additional hardware or compromising output performance, while seamlessly integrating with existing parallelism strategies and scheduling techniques. Specifically, LayerKV introduces layer-wise KV block allocation, management, and offloading for fine-grained control over system memory, coupled with an SLO-aware scheduler to optimize overall Service Level Objectives (SLOs). Comprehensive evaluations on representative models, ranging from 7B to 70B parameters, across various GPU configurations, demonstrate that LayerKV improves TTFT latency up to 69x and reduces SLO violation rates by 28.7%, significantly enhancing the user experience.

cs.DC↗