arXiv subjects

Hao Mark Chen

Publications and source records attributed to Hao Mark Chen.

At least 19 recordsLinked to original sources

arXiv2026

Planarian: Managing Agent State with Statepoints

LLM agents solve complex tasks by iteratively changing files, invoking local tools, and interacting with remote services, which modifies state across their local environment and remote services. Today, agents and users must manage these changes explicitly, whether reverting exploratory actions or recovering from erroneous ones. Doing so safely requires coordinated actions, yet current agent harnesses lack unified abstractions and mechanisms for managing local and remote state consistently and efficiently. We describe Planarian, an agent runtime with state management that enables agents and users to recover from erroneous actions and explore alternative executions over consistent local and remote environment state. Planarian introduces the abstraction of agent statepoints, which are consistent, restorable point-in-time versions of the environment state. Planarian exposes three state-management primitives to agents and users: (i) snapshot creates a new statepoint spanning local and remote state without requiring external services to support checkpoints: it relies on efficient incremental process and file system snapshotting to capture local sandboxed state, and transparently records compensating actions to undo remote state changes; (ii) rollback restores the environment to a previous statepoint by reverting to a prior local checkpoint and replaying compensating actions for remote state changes; and (iii) fork creates multiple isolated branches from a statepoint, enabling the agent to explore alternatives in parallel. We show that Planarian enables agents to undo mistakes and explore alternatives in parallel, improving task quality by up to 15x, and allows users to recover from erroneous actions with only 3% overhead.

cs.OS↗

arXiv2026

HM-ROUTER: Joint Model and Harness Routing for Agentic Systems

Agent performance depends on both the underlying model and the harness that manages its tool use and execution. Selecting a suitable pair requires accounting for their compatibility, yet training samples may cover only a subset of the growing combination space. We introduce HM-Router, a routing method that jointly selects a model and harness for each query. It learns separate model and harness representations shared across routes, with an interaction term inspired by canonical polyadic (CP) tensor decomposition to capture how their compatibility varies with the query. This sharing allows training samples from observed pairs to inform predictions for unobserved combinations. We curate a benchmark from 12 public agent benchmarks, covering 293 routes, 73 models, and 25 harnesses. HM-Router exceeds the strongest evaluated learned baseline by 7.3 percentage points in mean routing accuracy and leads at all seven evaluated cost budgets on the six-benchmark subset. When 90% of routes have their training outcomes withheld, allowing unobserved combinations improves normalized accuracy by 15.8 points over restricting the same router to observed routes. HM-Router has also demonstrated training sample efficiency for new routes and components and generalization to unseen benchmarks. Our code and data are open-sourced at https://github.com/hmarkc/HM-Router.

cs.LG↗

arXiv2026

HeteroReason: Heterogeneous FPGA-GPU Acceleration for Disaggregated Speculative Reasoning

Large Reasoning Models (LRMs) have achieved state-of-the-art performance in reasoning tasks by utilizing Chain-of-Thought (CoT) reasoning. To achieve fast execution speed, speculative reasoning techniques adopt a lightweight draft model for candidate token generation followed by process reward models (PRMs) for verification and a strong target model for refinements. This paper identifies that the existing speculative reasoning paradigm follows a strictly forward-only reasoning trajectory, which lacks robustness and can lead to severe error propagation if early reasoning steps are suboptimal. Furthermore, executing these disparate inference schemes, including sequential drafting and parallel verification on homogeneous GPU platforms, can lead to severe resource underutilization. To address this, we propose HeteroReason, an algorithm-hardware co-designed heterogeneous FPGA-GPU inference paradigm specifically tailored for LRM speculative reasoning. At the algorithmic level, we introduce a backtracking-enhanced workflow that enables the system to recover from low-quality states and explore alternative reasoning trajectories, significantly improving reasoning robustness. At the system level, the draft model is offloaded to the FPGA while deploying the PRM and target models on GPUs. A specialized workflow is optimized to achieve prefill-decode disaggregation, which exploits shadow synchronization to overlap GPU-side refinements with FPGA-side token updates to effectively hide synchronization latency. To mitigate inherent sequential constraints, we propose a step-ahead speculation and refinement scheduling scheme, transitioning the system from a sequential execution scheme to a parallel pipeline. Experimental evaluations show an average 4.2% accuracy improvement, with 1.01x-1.42x latency speedups and 1.25x-1.57x improvements in energy efficiency compared to homogeneous GPU baselines.

cs.AR↗

arXiv2026

HyQDB: LLM-Assisted Debugging for Hybrid Quantum Workflows

Hybrid quantum program failures frequently occur silently, yet existing debugging tools provide limited support for detecting and repairing them. These faults dominate the failures reported by domain experts, yet existing tools evaluate on public data that under-represents this failure mode. Our key insight is that faults divide into two classes that require different strategies: mechanical faults, which allow deterministic analysis, and conceptual faults, which require reconstructing the program's intent. To address this challenge, we present HyQDB, a tiered agent that injects deterministic hardware, physics and optimization evidence into the LLM repair process. When no evidence is detected, the agent treats this silence as a signal to escalate to a second intent-reconstruction tier, that infers the program's behaviour and reconciles it with the implementation. To evaluate HyQDB, we introduce QFaultBench, a benchmark built from an expert-derived fault taxonomy that represents the true failure modes of hybrid programs. On a held-out set of human-authored programs, HyQDB raises repair success over a standard LLM from 45% to 75%. We show that the escalation gate is crucial, giving a 20% improvement in mechanical fault repair accuracy.

cs.SE↗

arXiv2026

AgentKV: Phase-Aware KV Eviction for Agentic LLMs

Agentic serving can consume orders of magnitude more tokens than chatbot workloads, stressing both KV-cache capacity and decode-time bandwidth. Most KV-eviction methods score cached keys against representative queries drawn from the most recent tokens, assuming future attention resembles recent attention. We show that agentic generation violates this assumption: future queries form a mixture over think, act, tool, and others phases, and principal-angle analysis shows these components occupy measurably different query subspaces, so recency representatives systematically undervalue keys that upcoming phases will need. We propose AGENTKV, which maintains a small query buffer per phase and scores cached keys against their union. We further implement AGENTKV in a persistent multi-turn serving path that carries compressed KV state across turns and compacts retained KV pages online. Across two models, six task domains, and three KV budgets each, AGENTKV improves task score by 5.5 points on average over R-KV and 5.3 over Tri-attention. Relative to upstream full-KV SGLang, AGENTKV improves output-token throughput by up to 1.80x. Code: https://github.com/LiuTaowen-Tony/agentkv.

cs.LG↗

arXiv2026

DeepStack: Facilitating Co-Design Exploration of 3D DRAM-Stacked Accelerators for Distributed LLM Inference

Advances in hybrid bonding and packaging have driven growing interest in 3D DRAM-stacked AI accelerators. As large language models (LLMs) scale to hundreds of billions or trillions of parameters, distributed inference across multiple 3D chips has become essential for AI serving. This trend makes cross-stack co-design critical because system-level parallelization and scheduling choices are tightly coupled with hardware characteristics such as memory organization, interconnects, and thermal constraints. We present DeepStack, an accurate performance model and efficient design space exploration (DSE) framework for distributed 3D-stacked LLM inference. At the hardware level, DeepStack captures transaction-aware memory bandwidth, bank activation constraints, buffering limitations, and thermal and power behavior. At the system level, it incorporates comprehensive parallelization strategies and execution scheduling. Through a dual-stage network abstraction and tile-level compute-communication overlap modeling, DeepStack achieves up to 100,000x faster evaluation than state-of-the-art simulators at comparable accuracy. We cross-validate DeepStack against our in-house 3D designs, an NS-3 backend with 2.12% error, and vLLM serving on eight B200 GPUs with 12.92% error. Combined with hierarchical search, DeepStack efficiently explores about 2.5 x 10^14 design points spanning the number of stacked DRAM layers, DRAM vertical connectivity, interconnects, compute-memory allocation, and distributed scheduling under thermal and area constraints. A search-space ablation shows that restricted DSE baselines can miss up to 9.5x modeled throughput. Beyond modeling and DSE, DeepStack derives design implications for distributed 3D AI systems and guides performance optimization across the stack. Source code and artifacts are available at https://github.com/tile-ai/DeepStack/tree/ae.

cs.AR↗

arXiv2026

AcceptMoE: Commitment-Weighted Self-Sizing Verifier Expert Sets for Efficient MoE Speculative Decoding

Speculative decoding verifies a tree of draft tokens in one target-model forward pass. For a mixture-of-experts (MoE) target, however, parallel verification can activate the union of the experts selected by all tree nodes, even though only a small subset of those nodes reaches the accepted output. Token count, activated-expert union size, and expert-weight traffic are therefore distinct cost measures: reducing the token workload need not shrink the expert union proportionally, and under offloading, transfer traffic also depends on cache residency. We introduce AcceptMoE, a verifier-side expert selector that combines target-router scores with offline-estimated commitment probabilities and automatically adjusts the number of eligible experts for each verification block, eliminating the need for a user-specified expert budget. Under offloading, AcceptMoE conditions expert eligibility on cache residency instead of predicting natural routes and prefetching the corresponding expert weights. Although constraining target-expert eligibility changes the model distribution, across 12 model-task pairs spanning three MoE targets and four benchmarks, AcceptMoE's mean accuracy is 0.27 percentage points lower than that of EAGLE-3 speculative decoding with natural routing. Served with SGLang at batch size one, it reaches 1.290 times the throughput of this baseline with all expert weights in GPU memory, and 2.06 times under physical expert offloading, while reducing host-to-device traffic by 73.6 percent to 77.1 percent.

cs.LG↗

arXiv2026

AOSpec: Action and Observation Co-Speculation for Low-Latency Agent Serving

Large language model agents increasingly act through stateful tools, yet model generation and environment execution remain serialized at every step. As decoding accelerates, tool execution becomes a growing bottleneck. Existing action- or observation-only speculation leaves much of this latency exposed: value is concentrated in a few slow calls, some outcomes emerge only through execution, and longer lookahead typically requires an increasingly unlikely chain of action predictions. We present AOSpec, a lossless framework that co-speculates actions and observations across the full agent-environment loop. Expected Value Decoding (EVD) directs observation speculation toward outcomes with the greatest expected latency benefit, optimizing expected time hidden rather than hit rate. For outcomes only execution can reveal, AOSpec launches latency-critical target actions in isolated forks that contain their effects, while Joint Action-State Verification (JASV) verifies both the action and its origin state against committed execution before reuse. JASV recasts long-horizon action dependency from full-chain prediction into target action-state verification, breaking the lookahead--accuracy tradeoff and unlocking long-range overlap without sacrificing serial semantics. Across Terminal-Bench serving settings spanning four harnesses, five actor models, and five serving speeds, AOSpec outperforms every practical baseline, reducing mean end-to-end latency by 11.8-32.5% and p99 latency by up to 42.8%. Its gains increase as decoding accelerates, and its observation model transfers from Terminal-Bench to SWE-bench Verified without retraining.

cs.LG↗

arXiv2026

Model Guides You How to Draw: Adaptive Visual Gating for Unified Multimodal Reasoning

Unified multimodal models (UMMs) with interleaved reasoning, which generate both textual and visual steps as part of intermediate reasoning traces, have demonstrated great potential for visual mathematical reasoning tasks. However, we identify a key insight in this paradigm: generating intermediate visual reasoning steps is not always beneficial and can even be harmful, as self-generated visual steps may introduce erroneous visual evidence that misleads subsequent reasoning. Moreover, frequently triggering visual steps during reasoning incurs substantial computational and memory overhead, degrading inference efficiency. To address these accuracy and efficiency challenges, we observe that the model's internal signals can indicate whether a visual step will benefit reasoning before the entire visual generation is completed. Specifically, this work identifies two internal signals: 1) Generation Intent, which reflects whether the model has a concrete textual plan for what to draw, and 2) Visual Fidelity, which measures whether the visual generation remains grounded in the original input image. Leveraging these internal signals, we propose AdaViG, a training-free adaptive visual gating method for unified multimodal reasoning. AdaViG dynamically evaluates each triggered visual step at an early visual generation stage and aborts it when both signals are weak, thereby preventing misleading visual evidence from entering the reasoning trace while avoiding unnecessary computation. Comprehensive experiments demonstrate that AdaViG improves accuracy by up to 5.7% while reducing visual generation FLOPs by 25.0%-91.0% and wall-clock latency by 15.4%-45.6%.

cs.CV↗

arXiv2026

Context Memorization for Efficient Long Context Generation

Modern large language model (LLM) applications increasingly rely on long conditioning prefixes to control model behavior at inference time. While prefix-augmented inference is effective, it incurs two structural limitations: i) the prefix's influence fades as generation proceeds, and ii) attention computation over the prefix scales linearly with its length. Existing approaches either keep the prefix in attention while compressing it, or internalize it into model parameters through gradient-based training. The former still attends to the prefix at inference, while the latter is training-intensive and ill-suited to prefix updates. To address these issues, we propose attention-state memory, a training-free approach that externalizes the prefix into a lightweight, lookup-based memory of precomputed attention states between prefix and query tokens. On ManyICLBench with LLaMA-3.1-8B, our method improves accuracy over in-context learning at 1K-8K memory budgets while reducing attention latency by 1.36x at 8K, and surpasses full-attention RAG performance on NBA benchmark using only 20% of its memory footprint.

cs.CL↗

arXiv2026

AdaBlock-dLLM: Semantic-Aware Diffusion LLM Inference via Adaptive Block Size

Diffusion-based large language models (dLLMs) are gaining attention for their inherent capacity for parallel decoding, offering a compelling alternative to autoregressive LLMs. Among various decoding strategies, block-wise semi-autoregressive (semi-AR) approaches are widely adopted due to their support for KV caching and their favorable accuracy-speed trade-off. However, this paper identifies two fundamental limitations in the conventional semi-AR decoding approach that applies a fixed block size: i) late decoding overhead, where the unmasking of high-confidence tokens outside the current block is unnecessarily delayed, and ii) premature decoding error, where low-confidence tokens inside the current block are committed too early, leading to incorrect tokens. This paper presents the first systematic investigation challenging the fixed block size setting in semi-AR decoding. Through a statistical analysis of confidence dynamics during the denoising process, we identify a volatility band (VB) region during dLLM decoding, which encodes local semantic structure and can be used to guide adaptive block sizing. Leveraging these insights, we introduce AdaBlock-dLLM, a training-free, plug-and-play scheduler that adaptively aligns block boundaries with semantic steps by adjusting block size during runtime. Extensive experiments across diverse benchmarks show that AdaBlock-dLLM achieves up to 5.3% accuracy improvement under the same throughput budget. Beyond inference-time optimization, we hope our semantics-aware adaptive scheduling approach and confidence-based analysis will inspire future training strategies for dLLMs. Our code is available at https://github.com/lgxi24/AdaBlock-dLLM.

cs.LG↗

arXiv2026

FastTTS: Accelerating Test-Time Scaling for Edge LLM Reasoning

Recent advances in reasoning Large Language Models (LLMs) are driving the emergence of agentic AI systems. Edge deployment of LLM agents near end users is increasingly necessary to protect data privacy, enable offline use, and provide responsive interaction with local context. However, strict memory constraints on edge devices limit deployment to smaller LLMs, whose reasoning capabilities are much weaker than those of large cloud models, hindering practical deployment of edge agentic AI. Test-Time Scaling (TTS) offers a promising solution by allocating more compute during inference to enhance the reasoning capability of edge LLMs. However, current TTS methods introduce heavy hardware performance overhead on resource-constrained devices, making them impractical for real applications. To address this challenge, we present FastTTS, a serving system that enables fast and efficient TTS for memory-constrained LLM reasoning. After analyzing common patterns across various TTS methods and identifying their performance bottlenecks, we introduce three novel techniques: i) Speculative Beam Extension, which mitigates system stragglers caused by irregular reasoning paths, ii) Asymmetric Multi-Model Memory Allocation, which dynamically balances memory usage between token generation and reasoning-step verification, and iii) Dynamic Prefix-Aware Scheduling, which optimizes reasoning execution to maximize KV-cache reuse across search paths. FastTTS offers a plug-and-play third-party library on top of vLLM, enabling edge LLMs on a single consumer GPU to match cloud-model accuracy and cloud-measured latency. Comprehensive evaluation shows that FastTTS achieves an average 2.2x higher goodput and reduces latency by 38%--68% compared to the vLLM baseline; it pushes the boundaries of low-latency TTS on memory-constrained edge devices and highlights the potential for democratizing agentic AI.

cs.LG↗

arXiv2026

Dynamic Expert Sharing: Decoupling Memory from Parallelism in Mixture-of-Experts Diffusion LLMs

Among parallel decoding paradigms, diffusion large language models (dLLMs) have emerged as a promising candidate that balances generation quality and throughput. However, their integration with Mixture-of-Experts (MoE) architectures is constrained by an expert explosion: as the number of tokens generated in parallel increases, the number of distinct experts activated grows nearly linearly. This results in substantial memory traffic that pushes inference into a memory-bound regime, negating the efficiency gains of both MoE and parallel decoding. To address this challenge, we propose Dynamic Expert Sharing (DES), a novel technique that shifts MoE optimization from token-centric pruning and conventional expert skipping methods to sequence-level coreset selection. To maximize expert reuse, DES identifies a compact, high-utility set of experts to satisfy the requirements of an entire parallel decoding block. We introduce two innovative selection strategies: (1) Intra-Sequence Sharing (DES-Seq), which adapts optimal allocation to the sequence level, and (2) Saliency-Aware Voting (DES-Vote), a novel mechanism that allows tokens to collectively elect a coreset based on aggregated router weights. Extensive experiments on MoE dLLMs demonstrate that DES reduces unique expert activations by over 55% and latency by up to 38%, while retaining 99% of vanilla accuracy, effectively decoupling memory overhead from the degree of parallelism.

cs.LG↗

arXiv2025

Enhancing Trustworthiness with Mixed Precision: Benchmarks, Opportunities, and Challenges

Large language models (LLMs) have shown promising performance across various tasks. However, their autoregressive decoding process poses significant challenges for efficient deployment on existing AI hardware. Quantization alleviates memory and compute pressure by compressing weights, activations, and KV caches to low precisions while preserving generation quality. However, existing quantization frameworks typically focus on perplexity or classification accuracy, often omitting critical trustworthiness metrics. This gap introduces risks when applying quantized LLMs to downstream high-stakes domains such as finance and healthcare. In this work, we systematically investigate the impact of quantization on four trustworthiness metrics (adversarial robustness, fairness, machine ethics, and out-of-distribution robustness) and identify the instability across compression ratios and quantization methods. Building on these observations, we develop a novel precision-ensemble voting approach that leverages predictions from mixed-precision variants of the same model and consistently improves performance by up to $5.8\%$ on trustworthiness metrics. Our results highlight the importance of considering trustworthiness when developing model compression techniques and point to research opportunities at the intersection of compression and trustworthiness for safety-critical applications.

cs.LG↗

arXiv2025

Rethinking Optimal Verification Granularity for Compute-Efficient Test-Time Scaling

Test-time scaling (TTS) has proven effective in enhancing the reasoning capabilities of large language models (LLMs). Verification plays a key role in TTS, simultaneously influencing (1) reasoning performance and (2) compute efficiency, due to the quality and computational cost of verification. In this work, we challenge the conventional paradigms of verification, and make the first attempt toward systematically investigating the impact of verification granularity-that is, how frequently the verifier is invoked during generation, beyond verifying only the final output or individual generation steps. To this end, we introduce Variable Granularity Search (VG-Search), a unified algorithm that generalizes beam search and Best-of-N sampling via a tunable granularity parameter g. Extensive experiments with VG-Search under varying compute budgets, generator-verifier configurations, and task attributes reveal that dynamically selecting g can improve the compute efficiency and scaling behavior. Building on these findings, we propose adaptive VG-Search strategies that achieve accuracy gains of up to 3.1\% over Beam Search and 3.6\% over Best-of-N, while reducing FLOPs by over 52\%. We will open-source the code to support future research.

cs.AI↗

arXiv2025

Hardware-Aware Parallel Prompt Decoding for Memory-Efficient Acceleration of LLM Inference

The auto-regressive decoding of Large Language Models (LLMs) results in significant overheads in their hardware performance. While recent research has investigated various speculative decoding techniques for multi-token generation, these efforts have primarily focused on improving processing speed such as throughput. Crucially, they often neglect other metrics essential for real-life deployments, such as memory consumption and training cost. To overcome these limitations, we propose a novel parallel prompt decoding that requires only $0.0002$% trainable parameters, enabling efficient training on a single A100-40GB GPU in just 16 hours. Inspired by the human natural language generation process, $PPD$ approximates outputs generated at future timesteps in parallel by using multiple prompt tokens. This approach partially recovers the missing conditional dependency information necessary for multi-token generation, resulting in up to a 28% higher acceptance rate for long-range predictions. Furthermore, we present a hardware-aware dynamic sparse tree technique that adaptively optimizes this decoding scheme to fully leverage the computational capacities on different GPUs. Through extensive experiments across LLMs ranging from MobileLlama to Vicuna-13B on a wide range of benchmarks, our approach demonstrates up to 2.49$\times$ speedup and maintains a minimal runtime memory overhead of just $0.0004$%. More importantly, our parallel prompt decoding can serve as an orthogonal optimization for synergistic integration with existing speculative decoding, showing up to $1.22\times$ further speed improvement. Our code is available at https://github.com/hmarkc/parallel-prompt-decoding.

cs.LG↗

arXiv2025

FW-Merging: Scaling Model Merging with Frank-Wolfe Optimization

Model merging has emerged as a promising approach for multi-task learning (MTL), offering a data-efficient alternative to conventional fine-tuning. However, with the rapid development of the open-source AI ecosystem and the increasing availability of fine-tuned foundation models, existing model merging methods face two key limitations: (i) They are primarily designed for in-house fine-tuned models, making them less adaptable to diverse model sources with partially unknown model and task information, (ii) They struggle to scale effectively when merging numerous model checkpoints. To address these challenges, we formulate model merging as a constrained optimization problem and introduce a novel approach: Frank-Wolfe Merging (FW-Merging). Inspired by Frank-Wolfe optimization, our approach iteratively selects the most relevant model in the pool to minimize a linear approximation of the objective function and then executes a local merging similar to the Frank-Wolfe update. The objective function is designed to capture the desired behavior of the target-merged model, while the fine-tuned candidate models define the constraint set. More importantly, FW-Merging serves as an orthogonal technique for existing merging methods, seamlessly integrating with them to further enhance accuracy performance. Our experiments show that FW-Merging scales across diverse model sources, remaining stable with 16 irrelevant models and improving by 15.3% with 16 relevant models on 20 CV tasks, while maintaining constant memory overhead, unlike the linear overhead of data-informed merging methods. Compared with the state-of-the-art approaches, FW-Merging surpasses the data-free merging method by 32.8% and outperforms the data-informed Adamerging by 8.39% when merging 20 ViT models. Our code is open-sourced at github.com/hmarkc/FW-Merging.

cs.LG↗

arXiv2025

Enhancing LLM-based Quantum Code Generation with Multi-Agent Optimization and Quantum Error Correction

Multi-agent frameworks with Large Language Models (LLMs) have become promising tools for generating general-purpose programming languages using test-driven development, allowing developers to create more accurate and robust code. However, their potential has not been fully unleashed for domain-specific programming languages, where specific domain exhibits unique optimization opportunities for customized improvement. In this paper, we take the first step in exploring multi-agent code generation for quantum programs. By identifying the unique optimizations in quantum designs such as quantum error correction, we introduce a novel multi-agent framework tailored to generating accurate, fault-tolerant quantum code. Each agent in the framework focuses on distinct optimizations, iteratively refining the code using a semantic analyzer with multi-pass inference, alongside an error correction code decoder. We also examine the effectiveness of inference-time techniques, like Chain-of-Thought (CoT) and Retrieval-Augmented Generation (RAG) in the context of quantum programming, uncovering observations that are different from general-purpose code generation. To evaluate our approach, we develop a test suite to measure the impact each optimization has on the accuracy of the generated code. Our findings indicate that techniques such as structured CoT significantly improve the generation of quantum algorithms by up to 50%. In contrast, we have also found that certain techniques such as RAG show limited improvement, yielding an accuracy increase of only 4%. Moreover, we showcase examples of AI-assisted quantum error prediction and correction, demonstrating the effectiveness of our multi-agent framework in reducing the quantum errors of generated quantum programs.

quant-ph↗