LLM Inference

LLM: Large Language Model

Follow this topic Ask about this topic

Momentum

30 papers in the last four weeks, up 233% on the four weeks before. 0.3% of all new papers.

Jul 13Week of Sep 28

Latest papers 187

Jun 14, 2026cs.AR

Prefill/Decode-Aware Evaluation of LLM Inference on Emerging AI Accelerators

As large language models (LLMs) are increasingly deployed in latency- and cost-sensitive settings, inference efficiency has become a central systems challenge. While GPUs dominate current deployments, a growing number of AI accelerators claim advantages for LLM inference, yet it remains unclear under which conditions such accelerators outperform GPUs in practice. Recent inference systems decompose execution into Prefill and Decode phases, which exhibit distinct computational characteristics and latency metrics, commonly captured by time to first token (TTFT) and time per output token (TPOT). This paper presents a phase-aware evaluation of LLM inference performance across GPUs and emerging AI accelerators using a common model, Llama2-7B. By separately measuring Prefill and Decode performance, we reveal that accelerator advantages differ by phase and metric. Our results show that GPUs consistently excel in the compute-intensive Prefill phase, while GroqRack achieves significantly lower TPOT during Decode (batching not currently supported). However, GPUs regain an advantage in Decode throughput as batch size increases. These findings demonstrate that each platform exhibits distinct phase-dependent strengths. We further analyze heterogeneous Prefill/Decode disaggregation across different accelerator platforms, identifying performance gains and the workload and network conditions under which such gains are realized.

Shun Usami, Venkatram Vishwanath, E. Wes Bethel

Department of Computer Science San Francisco State University San Francisco, CA, 94132 · Argonne National Laboratory Lemont, IL, 60439 · Lawrence Berkeley National Laboratory Berkeley, CA, 94720

AI Accelerator Inference LLM Inference LLM Inference Serving

Jun 11, 2026cs.DC

STREAM: Multi-Tier LLM Inference Middleware with Dual-Channel HPC Token Streaming

Researchers and practitioners working with large language models face a fragmented landscape: local models are free and private but hardware limits the model size and context windows a researcher can use; institutional HPC centers offer powerful GPU resources at no marginal cost and keep data within institutional boundaries, but operate behind firewalls and are designed for batch jobs rather than interactive use; commercial cloud APIs provide frontier-model quality on demand but impose significant cost and data retention policies unsuitable for sensitive research data. No existing system unifies all three. STREAM (Smart Tiered Routing Engine for AI Models) addresses this gap with four contributions: (1) a three-tier routing architecture combining local, HPC, and cloud inference with a local LLM-based complexity judge; (2) a dual-channel HPC streaming architecture that separates the Globus Compute control plane (authentication and job dispatch) from a WebSocket relay data plane (token delivery), enabling sub-second TTFT (0.54 s median, 21.1x over batch mode's 11.40 s) through institutional firewalls without VPN or firewall rule changes, with end-to-end AES-256-GCM encryption ensuring the relay operator cannot read token payloads; (3) tier-aware context summarization that prevents long conversations from forcing simple queries onto expensive tiers; and (4) an HPC-as-API proxy mode that exposes HPC inference as an OpenAI-compatible endpoint callable from any standard client with no HPC expertise, a deployment pattern made practical only by the sub-second TTFT of contribution (2). Llama 3.2 3B achieves 85.1% free-tier retention on a 1,200-query benchmark spanning ten domains. Measured TTFT: 0.26 s local, 0.54 s HPC (relay), 1.68 s cloud.

Anas Nassar, Steve Mohr, Leonard Apanasevich +1

Advanced Cyberinfrastructure for Education and Research (ACER) University of Illinois Chicago Chicago, Illinois, USA

LLM Serving LLM Inference High-Performance Computing LLM Routing

Jun 10, 2026cs.SD

Real-Time Language Model Jamming: A Case Study for Live Music Accompaniment Generation

Language models (LMs) have become one of the most prominent paradigms in modern generative modeling. While making them faster has been the main focus of real-time deployment, speed alone is not enough. Many real-world applications, such as synchronized translation and voice synthesis, also require precise alignment between generation and external signals, both in terms of generation content and timing. We refer to this problem as \textit{frame-synchronous streaming inference}. To address it, we present StreamMUSE, an inference system that performs LM generation in response to an external signal stream within a client-server architecture. The client continuously sends high-frequency inference requests based on the most recent inputs and receives outputs synchronized to the external clock, while the server executes model inference. We demonstrate the framework through a live music accompaniment task, showing how real-time synchronization can be achieved across different deployment environments with varying round-trip latencies. We further model the relationship between system hyperparameters and round-trip latency, and evaluate how different environments affect optimal configurations to achieve real-time performance. Experimental results show a consistent correspondence between system real-time performance and music quality, demonstrating the effectiveness of the proposed framework. The project is open source. Relevant code and the latest updates are available at https://stream-muse-webpage.vercel.app/#audio-library.

Bowen Zheng, Andrew H. Yang, Jiaqi Ruan +5

University of Wisconsin–Madison, USA · University of California, San Diego, USA · University of Science and Technology of China, China +3

Jun 9, 2026cs.DC

TileFuse: A Fused Mixed-Precision Kernel Library for Efficient Quantized LLM Inference on AMD NPUs

With the growing demand for on-device LLM inference, edge SoCs increasingly integrate NPUs to improve performance and energy efficiency under tight power and thermal budgets. However, practical LLM deployment on current client NPUs remains difficult: widely used quantization formats such as AWQ do not map cleanly onto many existing NPU software stacks, which are often proprietary and expose limited low-level control. In this work, we present TileFuse, a close-to-metal mixed-precision kernel library for AMD XDNA2 NPUs that targets GEMM/GEMV-based operators in quantized LLM inference. TileFuse brings practical low-bit formats such as AWQ-style W4A16 and W8A16 directly onto XDNA2, rather than forcing the model to be reshaped around an NPU-specific quantization scheme. TileFuse co-designs weight layout, metadata placement, mixed-precision microkernels, and array-level dataflow. Specifically, it fuses unpacking, dequantization, and GEMM/GEMV execution into a single kernel flow, introduces an interleaved pre-tiling layout that supports GEMM dimensions up to 32K, and redesigns GEMV dataflow to utilize the full 4x8 AIE array. Across kernel-level evaluations, TileFuse improves performance by up to 121.6% for GEMM and 281% for GEMV over full-precision baselines, while delivering more than 2x performance and energy-efficiency gains over strong iGPU baselines on GEMM. In end-to-end LLM experiments on Ryzen AI laptops, TileFuse achieves up to 2.0x lower prefilling latency with more than 64.6% lower energy consumption. Together, these results show that XDNA2 is a practical target for AWQ-style edge LLM inference and that native NPU support for off-the-shelf quantization can make NPUs substantially more usable in real client deployments.

Wesley Pang, Gregory Hyegang Jun, Feiyang Liu +1

University of Illinois Urbana-Champaign Department of Electrical and Computer Engineering USA

Mixed-Precision Quantization LLM Quantization AI Accelerator Inference On-Device Language Model Inference LLM Inference

Jun 9, 2026cs.CL

Dep-LLM: Training-Free Depression Diagnosis via Evidence-Guided Structured Multi-factor with Reliable LLM Reasoning

Automatic Depression Detection (ADD) from clinical interviews is a pivotal task in computational mental health, yet it remains challenging due to two critical obstacles: 1) difficulty in modeling complex but sparsely distributed depression clues within lengthy, multi-topic clinical interviews, leading to superficial and unreliable reasoning; 2) scarcity of labeled data due to clinical privacy, together with high cost of training and fine-tuning, limiting the deployment of supervised ADD systems. To jointly address these challenges, we propose Dep-LLM, a training-free framework that mirrors the step-by-step reasoning of clinical psychiatrists and operates entirely on frozen off-the-shelf foundation LLMs. Dep-LLM comprises three stages. First, a Chain-of-Thought (CoT) Depression Multi-factor Analysis module structurally decomposes the long dialogue into five clinically aligned themes and produces evidence-grounded rationales, effectively handling long-context dependencies. Second, we introduce Confidence Analysis and Modulation module that quantifies the epistemic reliability from token-level entropy of each rationale and applies an intra-label and inter-theme modulation that amplifies trustworthy signals while suppressing uncertain ones without extra training. Third, a Collaborative Multi-factor Prediction module dynamically integrates multi-factor signals weighted by confidence into the final diagnosis. Extensive experiments on the DAIC-WOZ and E-DAIC datasets demonstrate the effectiveness and generalizability of Dep-LLM: it surpasses zero-shot baseline on nearly all 21 foundation LLMs across 9 metrics such as accuracy, macro F1 and weighted-average F1, and further outperforms state-of-the-art supervised domain-specific LLMs as well as the latest closed-source commercial LLMs, while requiring no extra training.

Yiqing Lyu, Xianbing Zhao, Buzhou Tang +1

LLM Inference Depression Detection Mental Health Clinical Reasoning Clinical NLP

Jun 8, 2026cs.CR

Steganography Without Modification: Hidden Communication via LLM Seeds

We demonstrate that widely deployed Large Language Model (LLM) inference stacks harbor a steganographic channel that requires no modification to model weights, sampling code, or output distributions. The channel exploits a structural property of deterministic decoding: pseudo-random number generators (PRNGs) used in inverse-transform sampling produce a seed-dependent sequence of token-level probability intervals that can be reconstructed from the generated text alone. A sender encodes a secret message in the PRNG seed before generation; a receiver reconstructs the intervals and recovers the seed, and thus the hidden payload, by exhaustive search over the seed space. We formalize two operational modes. In the known-prompt setting, sender and receiver share the prompt, enabling exact interval reconstruction and perfect seed recovery via forced alignment. In the unknown-prompt setting, only the generated text is available; approximate interval reconstruction combined with a maximum-hit-count scoring strategy still permits reliable recovery from sufficiently long outputs. Extensive experiments across six model families and five heterogeneous text domains show that, in the known-prompt setting, full 32-bit seed recovery from the complete 2^32 candidate space achieves up to 100% accuracy, depending on model and text domain, within 300 tokens and under 35 seconds on a single GPU. In the unknown-prompt setting, recovery reaches near-perfect accuracy at 600-800 tokens in about 12 seconds. We further analyze the influence of prompting strategies, tokenization ambiguities, and sampling hyperparameters on channel reliability. Moreover, we discuss several applications of our results: First, it allows for the steganographic transmission of 32 bits, but also shows that ignorance of the prompt is not a valid security assumption.

Felix Mächtle, Jonas Sander, Sebastian Berndt +3

Institute for IT Security, University of Lübeck, Lübeck, Germany · Technische Hochschule Lübeck, Lübeck, Germany

Language Model Decoding LLM Inference LLM Security

Jun 7, 2026cs.AI

Inference-Time Conformal Reasoning with Valid Factuality Control for Large Language Models

Large language models (LLMs) increasingly perform multi-step reasoning, where intermediate claims form implicit directed acyclic graphs whose node correctness is structurally conditioned on their ancestors. This makes factuality uncertainty structural, rather than a trivial accumulation of node-wise errors, and necessitates inference-time uncertainty quantification over the reasoning structure. While conformal prediction (CP) offers flexible user-specified factuality control, existing work remains post-hoc and cannot intervene during generation. To fill the gap between CP's flexibility and its post-hoc limitation, we propose an \emph{Inference-Time Conformal Reasoning (ITCR)} framework that integrates CP directly into reasoning graph generation. ITCR learns a structure-level factuality uncertainty function that aggregates claim-level factuality signals over reasoning graphs without complex modeling assumptions. We then design the non-conformity score based on graph-level factuality uncertainty and calibrate the conformal threshold to decide when to stop generation. We theoretically show such generation is nested, yielding valid coverage guarantees for factuality control. Experiments over multiple datasets and coverage objectives demonstrate empirically valid coverage. In downstream reasoning tasks, inference-time calibrated graphs yield more accurate generation than post-hoc pruned graphs.

Ting Wang, Yuanjie Shi, Yan Yan +1

University of Illinois Urbana-Champaign · School of EECS, Washington State University.

LLM Inference LLM Reasoning with Graphs Conformal Prediction LLM Uncertainty Estimation LLM Reasoning

Jun 6, 2026cs.AI

Cross-LLM Consistency in Inference: Evidence from Shared Interactions

Large language models (LLMs) differ in architecture, training data, and optimization procedures, yet they may still develop similar internal inference patterns. In this paper, we examine this hypothesis using interaction-based explanations. We find that LLMs often share interaction patterns when predicting the same target token from the same prompt. This consistency is more pronounced among advanced LLMs. Shared interactions also tend to be lower-order and show weaker positive-negative cancellation than non-shared interactions. These results suggest that advanced LLMs may be implicitly optimized toward common inference patterns, even though the mechanisms that give rise to such cross-model consistency remain open.

Siyu Lou, Yao Yan, Yuntian Chen +1

School of Computer Science Shanghai Jiao Tong University Ningbo Key Laboratory of Advanced Manufacturing Simulation Eastern Institute of Technology, Ningbo · College of Computer and Information Science Chongqing Normal University · Ningbo Key Laboratory of Advanced Manufacturing Simulation Eastern Institute of Technology, Ningbo

LLM Inference LLM Interpretability Cross-Lingual Consistency in Language Models

Jun 6, 2026cs.AI

When Does Delegation Beat Majority? A Delegation-Based Aggregator for Multi-Sample LLM Inference

Majority voting is the default unsupervised aggregator for multi-sample LLM inference, but it discards two signals: within-group answer entropy and between-group reasoning geometry. We aggregate by delegation instead (Propagational Proxy Voting, PPV): each group of samples keeps weight on its own answer in proportion to its entropy-based confidence (When) and routes the rest to peers by reasoning-embedding similarity (Whom); the stationary distribution of the resulting delegation matrix picks the consensus answer. This requires neither gold labels nor training. On MMLU-Pro with 128 samples per question, delegation beats majority by +1.5 pp overall and +2.24 pp on non-trivial questions (McNemar p ~ 1.0e-14, n = 8,099), overturning wrong majorities whose answer cluster is geometrically incoherent while the correct minority is tight. We then characterize exactly when delegation overturns majority: a two-option model gives a closed-form flip condition on each option's confidence and the weight it routes to the other, with a do-no-harm corollary for near-unanimous questions. The condition calls the realized winner on 96.5% of non-trivial questions, and its predicted mass gap tracks the realized gap at r = 0.97. We did not find any other unsupervised ensemble methods that close the oracle gap.

Yasushi Sakai, Allen Song, Kent Larson

1MIT Media Lab, Cambridge, MA

LLM Inference Language Model Ensembles Majority Voting

Jun 5, 2026cs.LG

Breaking the Ice: Analyzing Cold Start Latency in vLLM

As scalable inference services become popular, the cold start latency of an inference engine becomes important. Today, vLLM has evolved into the de-facto inference engine of choice for many inference workloads. Although popular, due to its complexity and rapid evolution, there has not been a systematic study on the startup latency of its engine. With major architectural innovations under it (e.g., the V1 API, introduction of torch.compile), in this paper, we present the first detailed performance characterization of vLLM startup latency. We break down the startup process into six foundational steps and demonstrate that this process is predominantly CPU-bound. Each step exhibits consistent and interpretable scaling trends with respect to model- and system-level parameters, enabling fine-grained attribution of latency sources. Building on these insights, we develop a lightweight analytical model that accurately predicts vLLM's startup latency for a given hardware configuration, providing actionable guidance for resource planning in large-scale inference environments. All our benchmarking datasets, analysis tools, and prediction scripts are open-sourced at https://github.com/upb-cn/vllm-startup-profiler

Huzaifa Shaaban Kabakibo, Animesh Trivedi, Lin Wang

Paderborn University, Paderborn, Germany · IBM Research Europe, Zurich, Switzerland

LLM Inference LLM Inference Serving

Jun 4, 2026cs.LG

Skip a Layer or Loop It? Learning Program-of-Layers in LLMs

Large language models (LLMs) perform inference by following a fixed depth and order, non-recurrent execution of all layers. We reveal the wide existence of training-free, flexible, dynamic program-of-layers (PoLar), where pretrained layers can be packed as modules and then skipped or looped to form a customized program for each input. For most inputs, substantially shorter program executions can achieve the same or better accuracy, while incorrect predictions of the original LLM can be corrected by alternative programs with fewer layers. These observations indicate that inference admits multiple valid latent computations beyond the standard forward pass. To efficiently achieve PoLar in practice, we propose a lightweight PoLar prediction network, which learns to generate execution programs that dynamically skip or repeat pretrained layers for each input. Experiments on mathematical reasoning benchmarks demonstrate that PoLar consistently improves accuracy over standard inference and prior dynamic-depth methods, often while executing fewer layers, and that these gains persist under out-of-distribution evaluation. Our results suggest that fixed-depth execution captures only a narrow subset of an LLM's latent reasoning capacity.

Ziyue Li, Yang Li, Tianyi Zhou

University of Maryland, College Park, MD, USA · 2MBZUAI, Abu Dhabi, UAE.

LLM Inference LLM Inference Acceleration LLM Reasoning

Jun 4, 2026cs.AI

Closing the Loop on Latent Reasoning via Test-Time Reconstruction

Recent work moves intermediate reasoning from natural-language traces into latent or cache-level representations to reduce token overhead and avoid a discrete communication bottleneck. However, this shift also removes a key advantage of textual reasoning: intermediate states are no longer inspectable, making it difficult to determine whether a latent state still preserves the constraints of the original query. As a result, latent reasoning typically operates in an open loop, where a latent state is produced and consumed without an input-anchored fidelity check. We propose ReLAT (Reconstruction-Guided Latent Reasoning At Test Time), a self-supervised test-time training method that closes this loop using the query itself as the reference. Our key observation is that if a latent state faithfully represents a query, the query should be recoverable from it; if the query cannot be recovered, the latent state has lost task-relevant information. ReLAT operationalizes this principle by constructing a differentiable Question -> Latent Thought -> Question cycle and optimizing query reconstruction loss through the latent thought before answer generation. This anchors opaque latent computation to the problem specification it is supposed to represent. Across mathematical reasoning, knowledge QA, and code generation benchmarks on the Qwen family, ReLAT consistently improves over single-model inference, text-based collaboration, open-loop latent collaboration, and alternative test-time training objectives. On Qwen3-8B, ReLAT raises AIME 2024 accuracy from 56.7% to 73.3%, a 16.6-point gain over the strongest open-loop latent baseline.

Xiaopeng Yuan, Haibo Jin, Ye Yu +4

University of Illinois Urbana-Champaign · Google · Florida State University

LLM Inference Test-Time Training LLM Reasoning Latent Reasoning

Jun 4, 2026cs.LG

Generative Criticality in Large Language Model Temperature Scaling

We propose a statistical-field framework for text generated by large language models (LLMs), treating token embeddings as continuous spin variables on a one-dimensional chain. Defining a susceptibility from the connected two-point correlator and an order parameter from the ensemble-averaged embedding field, we vary the \texttt{softmax} temperature

T

and observe a sharp susceptibility peak near a characteristic

T_c

with power-law-like scaling, a concurrent rapid change in the order parameter, and a collapse onto a single semantic direction below

T_c

. The intrinsic dimension estimated by the two nearest neighbor (TwoNN) method independently corroborates these findings, reaching a minimum near

T_c

. Results are robust across model scales (Qwen3: 0.6B--32B) and prompt categories. While the phenomenology closely resembles a continuous phase transition, the non-equilibrium nature of autoregressive generation warrants further investigation. Our framework provides quantitative tools for probing the collective statistical structure of LLM outputs and suggests connections between decoding strategies and critical phenomena.

Huajian Ruan, Jinyang Li, Xingyu Guo +1

State Key Laboratory of Nuclear Physics and Technology, Institute of Quantum Matter, South China Normal University, Guangzhou 510006, China · Key Laboratory of Atomic and Subatomic Structure and Quantum Control (MOE), Guangdong-Hong Kong Joint Laboratory of Quantum Matter, Guangzhou 510006, China · Guangdong Basic Research Center of Excellence for Structure and Fundamental Interactions of Matter, Guangdong Provincial Key Laboratory of Nuclear Science, Guangzhou 510006, China +4

LLM Inference Phase Transitions Representation Geometry in Language Models

Jun 4, 2026cs.CL

IR3DE: A Linear Router for Large Language Models

Foundational Large Language Models (LLMs) demonstrate proficiency on a wide range of general tasks, and achieve remarkable results on various specialized tasks via domain-expert LLMs. With the ever-growing list of available LLMs, inference routers are being proposed to select the most appropriate LLM for each prompt. However, existing routing methods either optimize cost across weak-to-strong generalist LLMs or require substantial training to support domain-expertise routing. In this paper, we propose IR3DE, a Ridge Regression-based Router for Domain Experts that provides cheap and fast routing decisions for each prompt. We evaluate IR3DE in two Causal Language Modeling (CLM) settings where the tasks are next-token prediction for all domains, and one reasoning setting where each domain has its own distinct reasoning task. Despite being a linear router, IR3DE achieves performance comparable to the other baselines in both CLM settings, and surpassing them in the reasoning setting, with a normalized performance of 98.4%. Moreover, IR3DE enables the addition or removal of new domain experts without requiring the router to be retrained from scratch, allowing a dynamic set of LLMs to be served with minimal disruption to the router itself. Our code is available at: github.com/gensyn-ai/IR3DE.

Eros Fanì, Oğuzhan Ersoy

Gensyn

LLM Inference LLM Routing Ridge Regression

Jun 4, 2026cs.LG

MolE-RAG: Molecular Structure-Enhanced Retrieval-Augmented Generation for Chemistry

Large language models (LLMs) have shown promise for molecular property prediction, but their ability to reason over chemical structures remains limited, as molecular representations such as SMILES differ substantially from the natural language on which LLMs are primarily trained. To bridge this semantic and chemical knowledge gap, we propose MolE-RAG, a training-free, molecule-centric retrieval-augmented generation framework for LLM-based molecular property prediction. MolE-RAG augments each prediction with three complementary sources of inference-time context: retrieved chemistry literature, molecule-specific information including compound synonyms, identifiers, functional group annotations, and physicochemical descriptors, and structurally similar molecules retrieved from the training set. We evaluate MolE-RAG across nine molecular property prediction tasks using proprietary, chemistry-specialized, and open-source LLMs. Across general-purpose LLMs, MolE-RAG improves ROC-AUC by up to 28 percentage points on classification tasks and reduces regression RMSE by up to 67% relative to a SMILES-only baseline. We further find that the utility of each context source varies across models and tasks, with different models benefiting most from textual retrieval, molecular context, or structural retrieval. These results suggest that molecule-centric retrieval can improve LLM-based molecular property prediction without model fine-tuning while providing a flexible framework for integrating heterogeneous chemical knowledge at inference time.

Joey Chan, Wonbin Kweon, Ashley Shin +4

University of Illinois Urbana-Champaign · University of California, San Diego

Retrieval-Augmented Generation LLM Inference Molecular Property Prediction

Jun 3, 2026cs.CR

SharedRequest: Privacy-Preserving Model-Agnostic Inference for Large Language Models

With the widespread deployment of public large language models (LLMs) such as ChatGPT, protecting user prompt privacy has become an increasingly critical issue. Existing privacy-preserving inference methods sacrifice either utility or efficiency, and often require model-specific modifications that limit their compatibility. In this paper, we propose SharedRequest, a model-agnostic framework for privacy-preserving LLM inference that reformulates privacy protection at the batch level rather than the individual-prompt level. The key idea is to obscure sensitive information by mixing original prompts with noisy variants, while grouping semantically equivalent instructions to amortize the inference cost over a large batch of queries with minimal impact on LLM response quality. This design is independent of the LLM architecture, requiring no access to model parameters or architectural modification. Empirical results demonstrate that SharedRequest achieves over

20\%

higher utility compared to prior differential privacy baselines, and its shared-prompt mechanism reduces query cost by up to

5\times

compared to non-batched inference.

Peihua Mai, Xuanrong Gao, Youlong Ding +3

1National University of Singapore (Chongqing) Research Institute · 2Chongqing Key Laboratory of Trusted Perception and Interaction Technology for Intelligent and Connected Vehicles, Chongqing, China · 3National University of Singapore +3

Privacy-Preserving Language Models LLM Inference Privacy Leakage in Language Models Efficient Language Model Inference

Jun 2, 2026cs.PF

NetKV: Network-Aware Decode Instance Selection for Disaggregated LLM Inference

Disaggregated LLM inference forces the KV cache to traverse the datacenter network before decoding begins, so transfer time enters directly into the Time to First Token (TTFT) budget. Current schedulers route on compute load and prefix-cache locality alone, ignoring the topological distance and dynamic congestion between prefill and decode instances. We close this gap with a thin operator-to-scheduler interface, the network cost oracle, and we prove that ignoring the network term renders cache-aware-only scheduling arbitrarily suboptimal as context length grows. NetKV, the O(|D|) per-request greedy that consumes this oracle, has tier rankings that are provably robust to stale telemetry. On a 64-GPU four-tier fat-tree simulator driven by Mooncake traces, NetKV reduces mean TTFT by up to 21.2% over round-robin and 17.6% over a tuned cache+load-aware scheduler, lifts SLO attainment by up to 20.1 percentage points, and keeps the Time Between Tokens overhead below 0.5 ms in every condition tested, with no changes to the transport, inference engine, or hardware.

Mubarak Adetunji Ojewale

Cloud Competency Centre, National College of Ireland

LLM Inference KV Caching LLM Inference Scheduling Cost-Aware Inference Distributed Inference

Jun 2, 2026cs.CL

Building Reliable Long-Form Generation via Hallucination Rejection Sampling

Large language models (LLMs) have achieved remarkable progress in open-ended text generation, yet they remain prone to hallucinating incorrect or unsupported content, which undermines their reliability. This issue is exacerbated in long-form generation due to hallucination snowballing, a phenomenon where early errors propagate and compound into subsequent outputs. To address this challenge, we propose a novel inference-time hallucination mitigation framework, named Segment-wise HAllucination Rejection Sampling (SHARS), which uses an arbitrary hallucination detector to identify and reject hallucinated segments during generation and resample until faithful content is produced. By retaining only confident information and building subsequent generations upon it, the framework mitigates hallucination accumulation and enhances factual consistency. To instantiate this framework, we adopt semantic uncertainty as the detector and introduce several vital modifications to address its limitations and better adapt it to long-form text. Our method enables models to self-correct hallucinations without requiring external resources such as web search or knowledge bases, while remaining compatible with them for future extensions. Empirical evaluations on standardized hallucination benchmarks demonstrate that our method substantially reduces hallucinations in long-form generation while preserving or even improving the informativeness of generation. Code is available at: https://github.com/TreeLLi/hallucination-rejection-sampling.

Lin Li, Georgia Channing, Suhaas M Bhat +2

Long-Form Document Generation LLM Hallucination Mitigation LLM Inference Rejection Sampling Text Generation

Jun 2, 2026cs.PF

DriftSched: Adaptive QoS-Aware Scheduling under Runtime Token Drift for Multi-Tenant GPU Inference

The rapid growth of large language model (LLM) inference services has increased the demand for efficient multi-tenant GPU scheduling. While modern inference runtimes such as vLLM improve throughput through continuous batching and optimized memory management, accurately estimating the runtime cost of heterogeneous inference requests remains challenging. In practice, admission-time workload estimates may deviate from observed execution behavior, leading to workload misclassification, queue imbalance, increased tail latency, and degraded Quality-of-Service (QoS). This paper presents DriftSched, a QoS-aware scheduling framework for multi-tenant LLM inference serving on NVIDIA L4 GPUs. DriftSched combines workload classification, token-budget estimation, tenant-aware queue management, and an online feedback mechanism to refine workload estimates using runtime observations. The framework evaluates FIFO, Priority, Weighted, Shortest-Job-First (SJF), and Aging Priority scheduling policies under heterogeneous multi-tenant workloads. Experimental results show that adaptive calibration reduces workload estimation error by an average of 38.8% (MAE) and 40.5% (RMSE), improving workload classification stability. Among all evaluated schedulers, SJF achieves the best overall performance, reducing median end-to-end latency by approximately 42% and P99 latency by approximately 16% relative to FIFO under sustained GPU contention. The results further indicate that scheduler selection has a greater impact on latency behavior than runtime calibration alone, while accurate workload characterization largely eliminates systematic estimation drift. This work contributes a reproducible framework for studying workload-estimation fidelity and QoS-aware scheduling in multi-tenant GPU inference systems.

Kathiravan Palaniappan

Independent Researcher (University of Colorado Colorado Springs Alumni)

LLM Inference LLM Inference Serving Inference-Time Compute Allocation

Jun 1, 2026cs.AR

Multi-Segment Attention: Enabling Efficient KV-Cache Management for Faster Large Language Model Serving

Large Language Model (LLM) inference relies on key-value (KV) caches to avoid redundant attention computation. While approximate KV cache retention techniques reduce memory usage by sacrificing model accuracy, lossless approaches instead evict KV cache blocks from GPU memory and reconstruct them on demand to preserve exact outputs. Existing lossless KV cache management systems primarily base eviction decisions on access frequency or positional heuristics, without considering how different KV cache blocks affect the execution efficiency of GPU attention kernels. In this paper, we propose AsymCache, a computation-latency-aware KV cache management system for LLM inference that explicitly aligns cache residency decisions with GPU attention kernel performance, including three key components: Multi-Segment Attention (MSA) for efficient non-contiguous KV context processing, a cache eviction policy that jointly optimizes hit rate and position-aware recomputation cost, and an adaptive chunking scheduler for high hardware utilization. Experiments show that AsymCache reduces TTFT by up to 1.90-2.03x and time-per-output-token (TPOT) by 1.62-1.71x over latest baselines, confirming the effectiveness of the method in common workloads and validating its design goal of balancing computational efficiency with cache hit rate. Moreover, the low-level design of AsymCache allows seamless integration into agent serving systems such as Continuum, where it further reduces average job latency by up to 18.1%.

Chunan Shi, Yilei Chen, Yilin Chen +2

Peking University Beijing, China

KV-Cache Offloading LLM Inference KV Caching KV-Cache Management LLM Inference Acceleration

Jun 1, 2026cs.DC

Not All Errors Are Equal: A Systematic Study of Error Propagation in Large Language Model Inference

Large language models (LLMs) are increasingly integrated into high-performance computing (HPC) workflows, accelerating scientific discovery through diverse perspectives such as code generation and domain-specific decision-making. Yet, how soft errors propagate and affect LLM inference remains largely unexplored. To bridge this gap, we present a comprehensive study on error propagation in LLM inference, enabled by our proposed LLMFI, a configurable and deterministic fault-injection framework. Using LLMFI, we systematically inject faults across three open-weighted LLMs and thirteen representative tasks, covering reasoning, multilingual, mathematical, and coding domains. In addition, we conduct fine-grained case studies that reveal critical vulnerability patterns. Overall, our study yields 17 takeaways that advance the understanding of error propagation in LLM inference and introduces four low-overhead directions to improve reliability through software-only modification, offering practical guidance for future error detection and mitigation.

Yafan Huang, Sheng Di, Guanpeng Li

University of Iowa · Argonne National Laboratory · University of Florida

LLM Inference Language Model Robustness

Jun 1, 2026cs.CL

Off-the-Shelf LLMs as Process Scorers: Training-Free Alternative to PRMs for Mathematical Reasoning

Selecting the best response from multiple small-model samples using a stronger scorer is a simple inference-time strategy, but fails when the small model has already committed to incorrect reasoning paths. PRM guided search avoids this by scoring candidate continuations during generation, but requires a reward model trained with step-level labels. We propose Chunk-Level Guided Generation, a training-free alternative that uses an off-the-shelf large language model as a process scorer. At each step, a small model samples k fixed-length candidate chunks, while the larger model scores the candidates using likelihoods without generating any text. The selected chunk is committed before the next step, steering generation before errors can propagate. We instantiate this framework with two selection rules: Likelihood-Guided Selection (LGS), which selects the chunk with the highest length-normalized large-model log-probability, and Contrastive-Guided Selection (CGS), which subtracts the small model's log-probability to favor chunks where the large model's preference diverges from the small model's. We show that scoring variable-length reasoning steps with large-model likelihoods is unreliable due to a systematic length bias that persists even after length normalization, and that fixed-length chunks avoid this confound. On GSM8K, MATH, Minerva Math, AMC23, and AIME24 with Qwen2.5-1.5B guided by Qwen2.5-32B and Llama-3.2-1B guided by Llama-3.1-70B, CGS outperforms majority voting by up to 28 pp and, under matched guidance budgets, matches or outperforms Qwen2.5-Math-PRM-72B guided search on most benchmarks without reward-model training. With Qwen2.5-7B guided by Qwen2.5-72B, CGS reaches 81.8% on MATH and 63.6% on Minerva Math at k=16, surpassing majority voting by 4--6 pp. Finally, Chunk-Level Guided Generation produces substantially shorter reasoning traces than PRM guided search.

Atoosa Chegini, Soheil Feizi

Department of Computer Science, University of Maryland

Numerical Reasoning in Language Models Inference-Time Search LLM Inference LLM Mathematical Reasoning

May 31, 2026cs.DC

Move the Query, Not the Cache: Characterizing Cross-Instance Latent Attention Redistribution Across GPU Fabrics

Frontier LLMs increasingly decide what a query attends to with a sparse-attention indexer that picks a few KV-cache blocks per query: attention's unit is now a small, reusable chunk. Agentic workloads hammer it: many sub-agents query one large codebase, reusing the same blocks. When that corpus outgrows one GPU it is partitioned across instances, so a query and the blocks it selects often sit on different GPUs: answering it means attention across instances. The reflex of prior cross-instance KV systems is to move the cache: pull the selected blocks to the requester. Multi-head Latent Attention inverts the arithmetic, compressing each token's key and value into one narrow vector, so a routed query row is only ~1 KB, smaller than the chunk it attends; routing the query is then often cheaper than moving the cache. Which primitive wins, over which fabric and request shape, is uncharted, least of all on device-initiated RDMA that makes per-request cross-node transfers cheap. We characterize cross-instance MLA attention on a real multi-node H100 cluster, distilling two reusable artifacts: a topology-aware cost model (probe / transfer / compute / return / merge) and a closed-form route/fetch/local predicate, whose constants we measure on real IBGDA, where the model tracks batched round-trips to within ~7%. At decode it routes the query, trading the cost of moving the cache (a ~3 ms re-adaptation splice for a contiguous chunk, or a scattered gather under selection) for a tens-of-microsecond round trip, and picks the fabric by probe latency, not peak bandwidth. We instantiate the cost model and predicate for MLA, but neither is MLA-specific: they apply wherever compression or sparse selection shrinks attention to small chunks (DeepSeek-V3.2, V4, and GLM-5.1 today). Extending them to a new architecture requires measuring just two coefficients: the routed payload and fetch's move-the-cache cost.

Bole Ma, Jan Eitzinger, Harald Köstler +1

LLM Inference KV Caching Sparse Attention Attention Mechanisms Distributed Inference Multi-Head Latent Attention

May 31, 2026cs.CL

Benchmarking Local LLMs for Natural-Language-to-SQL Querying in Biopharmaceutical Manufacturing: An Empirical Benchmark on Consumer-Grade Hardware

Biopharmaceutical manufacturing organizations operate under regulatory frameworks such as FDA guidance, EU Good Manufacturing Practice (GMP), and the EU AI Act, which can restrict the use of cloud-based artificial intelligence systems. Locally deployed large language models (LLMs) offer a privacy-preserving alternative, but their suitability for pharmaceutical manufacturing tasks remains underexplored. This study evaluates four open-source LLMs (Qwen 2.5 Coder 7B, Llama 3.1 8B, Mistral 7B, and Meditron 7B) deployed locally via Ollama for natural-language-to-SQL generation over a pharmaceutical manufacturing database. A FastAPI-based evaluation platform, PharmaBatchDB AI, was developed using a synthetic Microsoft SQL Server database containing approximately 63,000 records across Batch, Manufacturing Execution System (MES), and Clean-In-Place (CIP) modules. Models were benchmarked on 60 domain-specific natural-language questions using metrics including SQL extraction rate, SQL compliance, factual consistency, ROUGE-L, hallucination rate, throughput, and latency. Qwen 2.5 Coder 7B, Llama 3.1 8B, and Mistral 7B generated SQL for all evaluation tasks, while Meditron 7B failed on nearly all tasks due to context-window limitations and poor SQL generation capability. Llama 3.1 8B achieved the highest SQL compliance, whereas Qwen 2.5 Coder 7B achieved the strongest overall text similarity and factual consistency. Performance differences between the two leading models were not statistically significant. The results show that code-tuned general-purpose LLMs outperform a domain-specific biomedical model on structured query generation for pharmaceutical manufacturing data. Although fully local, GxP-aligned NLQ systems are feasible on consumer hardware, current performance levels still require human oversight and downstream validation for regulated use.

Sagar Bhetwal, Rajan Bastakoti, Nirajan Acharya +2

Department of Computer Science, University of the Cumberlands, Kentucky, United States · Department of Computer Science, DePaul University, Chicago, IL, United States · Youngstown State University, Youngstown, Ohio, United States +1

LLM Evaluation LLM Inference Text-to-SQL

May 30, 2026cs.CL

ProactiveLLM: Learning Active Interaction for Streaming Large Language Models

Standard Large Language Models (LLMs) follow a read-then-generate paradigm, causing unnecessary latency and computation. Streaming LLMs alleviate this issue by generating while receiving inputs, but still struggle to decide when to interact with the stream. Existing methods either hard-code interaction timing or rely on costly external alignment signals, such as timing labels, reasoning trajectories, or stronger teachers. In this paper, we propose ProactiveLLM, which achieves active interaction by leveraging the model's endogenous states to guide interaction decisions. The model first learns to perceive semantic sufficiency from partial inputs through two complementary training mechanisms: mask-based streaming modeling and synchronized privileged self-distillation (SPSD). The former applies monotonic random masking to the input during training, simulating progressively revealed streaming inputs and enabling the model to learn local semantic dependencies from partial-input views. The latter aligns the partial-context student view with a full-context teacher view generated by the same evolving model, allowing privileged full-context evidence to guide the student's understanding under incomplete observations. Together, these mechanisms induce endogenous sufficiency cues without requiring external teachers or annotations, providing a versatile foundation for the plug-and-play integration of diverse decision heads. Extensive evaluation across text and speech streaming tasks confirms that ProactiveLLM significantly reduces interaction latency while maintaining quality, validating its capacity for dynamic and active interaction. Code is publicly available at https://github.com/EIT-NLP/StreamingLLM/tree/main/ProactiveLLM.

Junlong Tong, Yao Zhang, Anhao Zhao +3

Eastern Institute of Technology, Ningbo · Shanghai Jiao Tong University · Hong Kong Polytechnic University +1

LLM Inference LLM Inference Acceleration

May 30, 2026cs.AI

Threshold-Based Exclusive Batching for LLM Inference

Mixed batching (MB)--interleaving prefill and decode in a single batch--has become the standard scheduling strategy for large language model (LLM) inference due to its efficiency in maximizing compute and memory utilization. However, through controlled experiments, we find that prefill-decode interference inflates MB's per-step marginal cost above that of pure decode. On the high-bandwidth H200 (4.8 TB/s), this occurs only when decode tokens exceed 80% of the batch; however, on the bandwidth-constrained RTX PRO 6000 (1.792 TB/s), this threshold plummets to just 20%. Consequently, the optimal choice between MB and exclusive batching (EB) fundamentally depends on GPU memory bandwidth, model size, and workload composition. We derive a closed-form condition for this EB-MB performance crossover, along with asymptotically optimal phase-switching thresholds and memory-safe batch sizing for EB. Optimized EB achieves up to 41.9% higher throughput on bandwidth-constrained GPUs, while MB retains its advantage on high-bandwidth hardware with larger models. Our hybrid scheduler EB+ applies this condition online to dynamically switch between EB and MB without manual intervention. Under non-stationary traffic with distribution or concurrency shifts, EB+ attains the highest or near-highest throughput in every setting, outperforming MB by up to 36.4%.

Weifang Zhang, Yuzhou Nie, Bowen Pang +2

The Hong Kong Polytechnic University · UC Santa Barbara · Technology and Engineering Center for Space Utilization, Chinese Academy of Sciences +1

GPU Acceleration LLM Inference LLM Inference Scheduling LLM Inference Acceleration

May 29, 2026cs.AI

Model-Native Computing Architecture: Envisioning Future System Architecture Through the Lens of Computer Architecture

Large language models are undergoing a transition from model technology to system technology. Engineering challenges like cache reuse, context capacity, agent scheduling, and permission control resemble classical computer systems problems. This raises a question: if we treat the LLM as a CPU, KV cache as processor cache, context window as main memory, and agent framework as an operating system, can decades of computer architecture wisdom guide next generation model native systems? This paper pursues this analogy as a visionary survey. We map computer architecture concepts onto the emerging model native stack, survey literature across LLM as OS, memory management, agent frameworks, tool protocols, multi agent coordination, cognitive architectures, and safety governance, finding that each addresses a different layer without a unifying model. We propose the Intelligent Computing Architecture (ICA): six functional layers with interface contracts and design axioms. We resolve the tension over whether the LLM resembles a CPU or OS via a dual plane architecture a probabilistic execution plane (what can be computed) and a deterministic control plane (what should be computed), with every layer passing through as a graded crossover. We propose three Amdahl style design heuristics Semantic Locality, Context Budget, and Agent Speedup as organizing back of envelope models, illustrate their parameter ranges with published data, and identify predictive validation as the principal open task. We articulate analogy boundaries, note differences between silicon and model era architectures, and propose a research roadmap. This is a conceptual and survey contribution with no new experimental results.

Hai Lin, Hoilam Pao, Shaoxiong Zhan +1

Shenzhen International Graduate School, Tsinghua University · Pengcheng Laboratory, Shenzhen, China

LLM Inference Persistent Memory for Language Models

May 29, 2026cs.CR

Bit-Exact AI Inference Verification Without Performance Tradeoffs

Verifying claims about AI workloads is a prerequisite for credible AI governance of covert adversaries (who comply with monitoring only when detection likelihood is high), yet the apparent non-determinism of GPU floating-point arithmetic forces auditors to accept approximate output matches. Covert adversaries can exploit unverifiable degrees of freedom in monitored computation. Attack vectors include steganography, unreported modification of inference software, and covert computation via unreported batch elements. Empirically, we analyze how modern inference engines (vLLM, HF transformers) produce deterministic but non-invariant outputs, without needing to set performance-compromising determinism flags, if the right information is available for re-computation and no atomic functions are called in the backend. We demonstrate that such bitwise-precise re-computation does not require access to identical hardware, via a software-only emulation of LLM inference across multiple NVIDIA GPU variants. Thus, accumulated rounding errors can be an auditable signature of the software and hardware setup used for inference, instead of a constraint on verifiability.

Naci Cankaya

Machine Intelligence Research Institute

Neural Network Verification LLM Inference LLM Auditing

May 29, 2026cs.LG

CoMem: Context Management with A Decoupled Long-Context Model

Context management enables agentic models to solve long-horizon tasks through iterative summarization of previous interaction histories. However, this process typically incurs substantial decoding overhead for the extra summarization tokens, which significantly affect the end-to-end response latency at deployment. In this paper, we introduce CoMem, a novel framework that decouples memory management from the primary agent workflow, enabling these processes to execute in parallel. We propose a

k

-step-off asynchronous pipeline that overlaps the memory model's summarization with the agent's inference, effectively masking the latency of context processing. To ensure robustness under this asynchronous setting, we introduce a reward-driven training strategy that aligns the memory model to capture sufficient statistics for the agent's decision-making. Theoretical analysis confirms that CoMem offers a superior efficiency-effectiveness trade-off compared to coupled architectures. Our extensive experimental results on SWE-Bench-Verified show that CoMem provides 1.4x latency improvements upon vanilla long-context solutions while preserving most of the performance. Furthermore, we demonstrate that these latency gains scale favorably with increased system throughput, offering a modular path forward for the independent optimization of agent reasoning and memory compression.

Yuwei Zhang, Chengyu Dong, Shuowei Jin +11

Halıcıoğlu Data Science Institute, University of California, San Diego · Amazon

LLM Inference LLM Inference Acceleration LLM Context Management Agent Memory Management

May 28, 2026cs.AR

Memory-Bound but Not Bandwidth-Limited: The Physical AI Inference Gap in Batch-1 LLM Decode

Physical AI systems, including robots, autonomous vehicles, embodied agents and edge copilots, often run a different inference workload from cloud LLM serving: single-stream, batch-1 autoregressive decode, where one robot, camera feed or user session waits on the next token. This workload is usually described as memory-bandwidth-bound. Each decode step streams model weights and the active KV cache, so latency should scale with peak HBM bandwidth. We show that this account is true but incomplete. We measure batch-1 decode for three 7 to 8B-class GQA transformers across four NVIDIA GPUs: H100 SXM5, A100-80GB SXM4, L40S and L4. We evaluate context lengths from 2048 to 16384, producing 44 valid cells under a controlled bf16 SDPA setup. The achieved fraction of peak HBM bandwidth falls as peak bandwidth rises. On the headline Qwen-2.5-7B ctx=2048 cell, an L4 reaches roughly 81 percent of its analytic memory floor, while an H100 reaches only 27 percent. Physical-AI decode is memory-dominated, but faster memory does not translate into proportional latency gains. We test the missing term with a CUDA Graphs A/B experiment. On H100 at ctx=2048, CUDA Graphs improves decode latency by 1.259x across N=10 fresh sessions, with a 95 percent bootstrap confidence interval of 1.253 to 1.267. On L4, the same intervention gives only 1.028x. This isolates a launch-side overhead that becomes visible on fast GPUs but remains mostly hidden on slower, bandwidth-bound GPUs. The deployment implication is that memory savings matter only when the runtime realises them. On L4, bf16 decode sits close to the memory floor, but common quantised paths do not recover the expected 4x weight-traffic reduction: bnb-nf4 reaches 59.36 ms/step and AutoAWQ+Marlin reaches 45.24 ms/step from a 62.32 ms bf16 baseline. GPTQ+ExLlamaV2, with Ada-tuned int4 kernels, reaches 17.36 ms/step.

Josef Chen

KAIKAKU

GPU Kernel Optimization GPU Acceleration LLM Inference