Daily Papers

We introduce PHLoRA (Pronounced "flora"). (Post-hoc LoRA), a simple yet powerful method to extract low-rank adaptation adapters from full-rank fine-tuned models without requiring access to training data or gradients. By computing the low-rank decomposition of weight differences between a base model and its fine-tuned counterpart, our method reconstructs adapter modules that can be merged or dynamically routed at inference time via S-LoRA, or served in scalable, industry settings using platforms like NVIDIA NIM. This approach amortizes latency overhead across requests and yields substantial cost savings. Unlike prior work that trains each adapter explicitly, our approach decouples fine-tuning from adapter generation, allowing adapter extraction from existing full-rank models or third-party checkpoints. Experiments on text, image, and video benchmarks using the Amazon Nova model family demonstrate that extracted adapters preserve high energy from the full weight delta, can be pruned safely, and yield negligible degradation in downstream task performance when re-merged. Overall, PHLoRA provides a practical path for making all existing full-rank checkpoints adapter-ready, democratizing scalable inference for all models.

The A2AS framework is introduced as a security layer for AI agents and LLM-powered applications, similar to how HTTPS secures HTTP. A2AS enforces certified behavior, activates model self-defense, and ensures context window integrity. It defines security boundaries, authenticates prompts, applies security rules and custom policies, and controls agentic behavior, enabling a defense-in-depth strategy. The A2AS framework avoids latency overhead, external dependencies, architectural changes, model retraining, and operational complexity. The BASIC security model is introduced as the A2AS foundation: (B) Behavior certificates enable behavior enforcement, (A) Authenticated prompts enable context window integrity, (S) Security boundaries enable untrusted input isolation, (I) In-context defenses enable secure model reasoning, (C) Codified policies enable application-specific rules. This first paper in the series introduces the BASIC security model and the A2AS framework, exploring their potential toward establishing the A2AS industry standard.

Generative retrieval has emerged as a powerful paradigm for LLM-based recommendation. However, industrial recommender systems often benefit from restricting the output space to a constrained subset of items based on business logic (e.g. enforcing content freshness or product category), which standard autoregressive decoding cannot natively support. Moreover, existing constrained decoding methods that make use of prefix trees (Tries) incur severe latency penalties on hardware accelerators (TPUs/GPUs). In this work, we introduce STATIC (Sparse Transition Matrix-Accelerated Trie Index for Constrained Decoding), an efficient and scalable constrained decoding technique designed specifically for high-throughput LLM-based generative retrieval on TPUs/GPUs. By flattening the prefix tree into a static Compressed Sparse Row (CSR) matrix, we transform irregular tree traversals into fully vectorized sparse matrix operations, unlocking massive efficiency gains on hardware accelerators. We deploy STATIC on a large-scale industrial video recommendation platform serving billions of users. STATIC produces significant product metric impact with minimal latency overhead (0.033 ms per step and 0.25% of inference time), achieving a 948x speedup over a CPU trie implementation and a 47-1033x speedup over a hardware-accelerated binary-search baseline. Furthermore, the runtime overhead of STATIC remains extremely low across a wide range of practical configurations. To the best of our knowledge, STATIC enables the first production-scale deployment of strictly constrained generative retrieval. In addition, evaluation on academic benchmarks demonstrates that STATIC can considerably improve cold-start performance for generative retrieval. Our code is available at https://github.com/youtube/static-constraint-decoding.

Recent years have witnessed remarkable advances in talking head generation, owing to its potential to revolutionize the human-AI interaction from text interfaces into realistic video chats. However, research on text-driven talking heads remains underexplored, with existing methods predominantly adopting a cascaded pipeline that combines TTS systems with audio-driven talking head models. This conventional pipeline not only introduces system complexity and latency overhead but also fundamentally suffers from asynchronous audiovisual output and stylistic discrepancies between generated speech and visual expressions. To address these limitations, we introduce OmniTalker, an end-to-end unified framework that simultaneously generates synchronized speech and talking head videos from text and reference video in real-time zero-shot scenarios, while preserving both speech style and facial styles. The framework employs a dual-branch diffusion transformer architecture: the audio branch synthesizes mel-spectrograms from text, while the visual branch predicts fine-grained head poses and facial dynamics. To bridge modalities, we introduce a novel audio-visual fusion module that integrates cross-modal information to ensure temporal synchronization and stylistic coherence between audio and visual outputs. Furthermore, our in-context reference learning module effectively captures both speech and facial style characteristics from a single reference video without introducing an extra style extracting module. To the best of our knowledge, OmniTalker presents the first unified framework that jointly models speech style and facial style in a zero-shot setting, achieving real-time inference speed of 25 FPS. Extensive experiments demonstrate that our method surpasses existing approaches in generation quality, particularly excelling in style preservation and audio-video synchronization.

Deployed language models must produce outputs that are both correct and format-compliant. We study this structured-output reliability gap using two mathematical benchmarks -- GSM8K and MATH -- as a controlled testbed: ground truth is unambiguous and the output contract is strict (JSON with required fields). We evaluate three 7-9B models under five prompting strategies and report output accuracy -- the joint event of mathematical correctness and valid JSON structure -- as the primary metric. A systematic format failure emerges: NAIVE prompting (no system prompt) achieves up to 85% task accuracy on GSM8K but 0% output accuracy across all models and datasets. REFERENCE prompting (a minimal hand-written JSON format prompt) fares little better, yielding 0% output accuracy for two of four models tested. Constrained decoding enforces syntactic validity but incurs 3.6x-8.2x latency overhead and in several settings degrades task performance substantially. To overcome this limitation, we developed AloLab, an iterative system-prompt optimizer (meta-agent: Claude Sonnet 4.5) requiring only black-box API access to the target model; it reaches 84-87% output accuracy on GSM8K and 34-40% on MATH across five independent runs per model, with 29/30 paired McNemar comparisons against the best static prompt significant at p < 0.05, at near-NAIVE inference latency and without model fine-tuning. The same format failure extends to GPT-4o (OpenAI, 2024), a proprietary closed-source model: REFERENCE achieves 0% output accuracy due to systematic markdown-fence wrapping, while AloLab reaches 95.2% [94.8, 95.6]. An ablation replacing the Sonnet 4.5 meta-agent with Claude 3 Haiku reduces mean output accuracy to 61.0% and increases run-to-run standard deviation from <1 pp to 21.8 pp, confirming that meta-agent capability is a primary driver of optimization quality.

As AI agents become widely deployed as online services, users often rely on an agent developer's claim about how safety is enforced, which introduces a threat where safety measures are falsely advertised. To address the threat, we propose proof-of-guardrail, a system that enables developers to provide cryptographic proof that a response is generated after a specific open-source guardrail. To generate proof, the developer runs the agent and guardrail inside a Trusted Execution Environment (TEE), which produces a TEE-signed attestation of guardrail code execution verifiable by any user offline. We implement proof-of-guardrail for OpenClaw agents and evaluate latency overhead and deployment cost. Proof-of-guardrail ensures integrity of guardrail execution while keeping the developer's agent private, but we also highlight a risk of deception about safety, for example, when malicious developers actively jailbreak the guardrail. Code and demo video: https://github.com/SaharaLabsAI/Verifiable-ClawGuard

Diffusion models are proficient at generating high-quality images. They are however effective only when operating at the resolution used during training. Inference at a scaled resolution leads to repetitive patterns and structural distortions. Retraining at higher resolutions quickly becomes prohibitive. Thus, methods enabling pre-existing diffusion models to operate at flexible test-time resolutions are highly desirable. Previous works suffer from frequent artifacts and often introduce large latency overheads. We propose two simple modules that combine to solve these issues. We introduce a Frequency Modulation (FM) module that leverages the Fourier domain to improve the global structure consistency, and an Attention Modulation (AM) module which improves the consistency of local texture patterns, a problem largely ignored in prior works. Our method, coined Fam diffusion, can seamlessly integrate into any latent diffusion model and requires no additional training. Extensive qualitative results highlight the effectiveness of our method in addressing structural and local artifacts, while quantitative results show state-of-the-art performance. Also, our method avoids redundant inference tricks for improved consistency such as patch-based or progressive generation, leading to negligible latency overheads.

The growing demand for long-context inference capabilities in Large Language Models (LLMs) has intensified the computational and memory bottlenecks inherent to the standard attention mechanism. To address this challenge, we introduce BLASST, a drop-in sparse attention method that dynamically prunes the attention matrix without any pre-computation or proxy scores. Our method uses a fixed threshold and existing information from online softmax to identify negligible attention scores, skipping softmax computation, Value block loading, and the subsequent matrix multiplication. This fits seamlessly into existing FlashAttention kernel designs with negligible latency overhead. The approach is applicable to both prefill and decode stages across all attention variants (MHA, GQA, MQA, and MLA), providing a unified solution for accelerating long-context inference. We develop an automated calibration procedure that reveals a simple inverse relationship between optimal threshold and context length, enabling robust deployment across diverse scenarios. Maintaining high accuracy, we demonstrate a 1.62x speedup for prefill at 74.7% sparsity and a 1.48x speedup for decode at 73.2% sparsity on modern GPUs. Furthermore, we explore sparsity-aware training as a natural extension, showing that models can be trained to be inherently more robust to sparse attention patterns, pushing the accuracy-sparsity frontier even further.

Large language models (LLMs) excel across diverse tasks but face significant deployment challenges due to high inference costs. LLM inference comprises prefill (compute-bound) and decode (memory-bound) stages, with decode dominating latency particularly for long sequences. Current decoder-only models handle both stages uniformly, despite their distinct computational profiles. We propose OverFill, which decouples these stages to optimize accuracy-efficiency tradeoffs. OverFill begins with a full model for prefill, processing system and user inputs in parallel. It then switches to a dense pruned model, while generating tokens sequentially. Leveraging more compute during prefill, OverFill improves generation quality with minimal latency overhead. Our 3B-to-1B OverFill configuration outperforms 1B pruned models by 83.2%, while the 8B-to-3B configuration improves over 3B pruned models by 79.2% on average across standard benchmarks. OverFill matches the performance of same-sized models trained from scratch, while using significantly less training data. Our code is available at https://github.com/friendshipkim/overfill.

Modern Machine Learning models are expensive IP and business competitiveness often depends on keeping this IP confidential. This in turn restricts how these models are deployed -- for example it is unclear how to deploy a model on-device without inevitably leaking the underlying model. At the same time, confidential computing technologies such as Multi-Party Computation or Homomorphic encryption remain impractical for wide adoption. In this paper we take a different approach and investigate feasibility of ML-specific mechanisms that deter unauthorized model use by restricting the model to only be usable on specific hardware, making adoption on unauthorized hardware inconvenient. That way, even if IP is compromised, it cannot be trivially used without specialised hardware or major model adjustment. In a sense, we seek to enable cheap locking of machine learning models into specific hardware. We demonstrate that locking mechanisms are feasible by either targeting efficiency of model representations, such making models incompatible with quantisation, or tie the model's operation on specific characteristics of hardware, such as number of cycles for arithmetic operations. We demonstrate that locking comes with negligible work and latency overheads, while significantly restricting usability of the resultant model on unauthorized hardware.

Computer-use agents (CUAs) have made tremendous progress in the past year, yet they still frequently produce misaligned actions that deviate from the user's original intent. Such misaligned actions may arise from external attacks (e.g., indirect prompt injection) or from internal limitations (e.g., erroneous reasoning). They not only expose CUAs to safety risks, but also degrade task efficiency and reliability. This work makes the first effort to define and study misaligned action detection in CUAs, with comprehensive coverage of both externally induced and internally arising misaligned actions. We further identify three common categories in real-world CUA deployment and construct MisActBench, a benchmark of realistic trajectories with human-annotated, action-level alignment labels. Moreover, we propose DeAction, a practical and universal guardrail that detects misaligned actions before execution and iteratively corrects them through structured feedback. DeAction outperforms all existing baselines across offline and online evaluations with moderate latency overhead: (1) On MisActBench, it outperforms baselines by over 15% absolute in F1 score; (2) In online evaluation, it reduces attack success rate by over 90% under adversarial settings while preserving or even improving task success rate in benign environments.

Decentralized Autonomous Organizations (DAOs) are inclined explore Small Language Models (SLMs) as edge-native constitutional firewalls to vet proposals and mitigate semantic social engineering. While scaling inference-time compute (System 2) enhances formal logic, its efficacy in highly adversarial, cryptoeconomic governance environments remains underexplored. To address this, we introduce Sentinel-Bench, an 840-inference empirical framework executing a strict intra-model ablation on Qwen-3.5-9B. By toggling latent reasoning across frozen weights, we isolate the impact of inference-time compute against an adversarial Optimism DAO dataset. Our findings reveal a severe compute-accuracy inversion. The autoregressive baseline (System 1) achieved 100% adversarial robustness, 100% juridical consistency, and state finality in under 13 seconds. Conversely, System 2 reasoning introduced catastrophic instability, fundamentally driven by a 26.7% Reasoning Non-Convergence (cognitive collapse) rate. This collapse degraded trial-to-trial consensus stability to 72.6% and imposed a 17x latency overhead, introducing critical vulnerabilities to Governance Extractable Value (GEV) and hardware centralization. While rare (1.5% of adversarial trials), we empirically captured "Reasoning-Induced Sycophancy," where the model generated significantly longer internal monologues (averaging 25,750 characters) to rationalize failing the adversarial trap. We conclude that for edge-native SLMs operating under Byzantine Fault Tolerance (BFT) constraints, System 1 parameterized intuition is structurally and economically superior to System 2 iterative deliberation for decentralized consensus. Code and Dataset: https://github.com/smarizvi110/sentinel-bench

Private large language model (LLM) inference based on secure multi-party computation (MPC) offers cryptographically-secure protection for both user prompt and proprietary model weights. However, it suffers from large latency overhead especially for long input sequences. While key-value (KV) cache eviction algorithms have been proposed to reduce the computation and memory cost for plaintext inference, they are not designed for MPC and cannot benefit private inference easily. In this paper, we propose an accurate and MPC-friendly KV cache eviction framework, dubbed MPCache. MPCache is built on the observation that historical tokens in a long sequence may have different effects on the downstream decoding. Hence, MPCache combines a look-once static eviction algorithm to discard unimportant tokens and a query-aware dynamic selection algorithm to further select a small subset of tokens for attention computation. As existing dynamic selection algorithms incur too much latency, we propose a series of optimizations to drastically reduce the KV cache selection overhead, including MPC-friendly similarity approximation, hierarchical KV cache clustering, and cross-layer index sharing strategy. With extensive experiments, we demonstrate that MPCache consistently outperforms prior-art KV cache eviction baselines across different LLM generation tasks and achieves 1.8~2.01x and 3.39~8.37x decoding latency and communication reduction on different sequence lengths, respectively.

Contrastive pretraining of image-text foundation models, such as CLIP, demonstrated excellent zero-shot performance and improved robustness on a wide range of downstream tasks. However, these models utilize large transformer-based encoders with significant memory and latency overhead which pose challenges for deployment on mobile devices. In this work, we introduce MobileCLIP -- a new family of efficient image-text models optimized for runtime performance along with a novel and efficient training approach, namely multi-modal reinforced training. The proposed training approach leverages knowledge transfer from an image captioning model and an ensemble of strong CLIP encoders to improve the accuracy of efficient models. Our approach avoids train-time compute overhead by storing the additional knowledge in a reinforced dataset. MobileCLIP sets a new state-of-the-art latency-accuracy tradeoff for zero-shot classification and retrieval tasks on several datasets. Our MobileCLIP-S2 variant is 2.3times faster while more accurate compared to previous best CLIP model based on ViT-B/16. We further demonstrate the effectiveness of our multi-modal reinforced training by training a CLIP model based on ViT-B/16 image backbone and achieving +2.9% average performance improvement on 38 evaluation benchmarks compared to the previous best. Moreover, we show that the proposed approach achieves 10times-1000times improved learning efficiency when compared with non-reinforced CLIP training.

We describe an adaptation of VACE (Video All-in-one Creation and Editing) for real-time autoregressive video generation. VACE provides unified video control (reference guidance, structural conditioning, inpainting, and temporal extension) but assumes bidirectional attention over full sequences, making it incompatible with streaming pipelines that require fixed chunk sizes and causal attention. The key modification moves reference frames from the diffusion latent space into a parallel conditioning pathway, preserving the fixed chunk sizes and KV caching that autoregressive models require. This adaptation reuses existing pretrained VACE weights without additional training. Across 1.3B and 14B model scales, VACE adds 20-30% latency overhead for structural control and inpainting, with negligible VRAM cost relative to the base model. Reference-to-video fidelity is severely degraded compared to batch VACE due to causal attention constraints. A reference implementation is available at https://github.com/daydreamlive/scope.

Key-value (KV) cache compression has emerged as a critical technique for reducing the memory and latency overhead of autoregressive language models during inference. Prior approaches predominantly rely on query-key attention scores to rank and evict cached tokens, assuming that attention intensity correlates with semantic importance. However, this heuristic overlooks the contribution of value vectors, which directly influence the attention output. In this paper, we propose CurDKV, a novel, value-centric KV compression method that selects keys and values based on leverage scores computed from CUR matrix decomposition. Our approach approximates the dominant subspace of the attention output softmax(QK^T)V, ensuring that the retained tokens best preserve the model's predictive behavior. Theoretically, we show that attention score approximation does not guarantee output preservation, and demonstrate that CUR-based selection minimizes end-to-end attention reconstruction loss. Empirically, CurDKV achieves up to 9.6% higher accuracy than state-of-the-art methods like SnapKV and ChunkKV under aggressive compression budgets on LLaMA and Mistral, while maintaining compatibility with FlashAttention and Grouped Query Attention. In addition to improved accuracy, CurDKV reduces generation latency by up to 40% at high compression, offering a practical speed-accuracy tradeoff.

Expert-Specialized Fine-Tuning (ESFT) adapts Mixture-of-Experts (MoE) large language models to enhance their task-specific performance by selectively tuning the top-activated experts for the task. Serving these fine-tuned models at scale is challenging: deploying merged models in isolation is prohibitively resource-hungry, while existing multi-adapter serving systems with LoRA-style additive updates are incompatible with ESFT's expert-oriented paradigm. We present ExpertWeave, a system that serves multiple ESFT adapters concurrently over a single shared MoE base model, drastically reducing the memory footprint and improving resource utilization. To seamlessly integrate into existing inference pipelines for MoE models with non-intrusive modifications and minimal latency overhead, ExpertWeave introduces a virtual-memory-assisted expert weight manager that co-locates base-model and adapter experts without incurring memory overhead from fragmentation, and a fused kernel for batched rerouting to enable lightweight redirection of tokens to the appropriate experts at runtime. Our evaluations show that ExpertWeave can simultaneously serve multiple adapters of a 16B MoE model on a single accelerator where the baseline runs out of memory, or provides up to 94x more KV cache capacity and achieves up to 18% higher throughput while using comparable resources, all without compromising model accuracy. ExpertWeave maintains low overhead even when scaling to 20 adapters, with a 4-11% latency increase compared with serving the base model alone. Source code will be released soon.

Prior parameter-efficient fine-tuning (PEFT) algorithms reduce memory usage and computational costs of fine-tuning large neural network models by training only a few additional adapter parameters, rather than the entire model. However, the reduction in computational costs due to PEFT does not necessarily translate to a reduction in training time; although the computational costs of the adapter layers are much smaller than the pretrained layers, it is well known that those two types of layers are processed sequentially on GPUs, resulting in significant latency overhead. LoRA and its variants merge low-rank adapter matrices with pretrained weights during inference to avoid latency overhead, but during training, the pretrained weights remain frozen while the adapter matrices are continuously updated, preventing such merging. To mitigate this issue, we propose Partial Connection Adaptation (PaCA), which fine-tunes randomly selected partial connections within the pretrained weights instead of introducing adapter layers in the model. PaCA not only enhances training speed by eliminating the time overhead due to the sequential processing of the adapter and pretrained layers but also reduces activation memory since only partial activations, rather than full activations, need to be stored for gradient computation. Compared to LoRA, PaCA reduces training time by 22% and total memory usage by 16%, while maintaining comparable accuracy across various fine-tuning scenarios, such as fine-tuning on the MMLU dataset and instruction tuning on the Oasst1 dataset. PaCA can also be combined with quantization, enabling the fine-tuning of large models such as LLaMA3.1-70B. In addition, PaCA enables training with 23% longer sequence and improves throughput by 16% on both NVIDIA A100 GPU and INTEL Gaudi2 HPU compared to LoRA. The code is available at https://github.com/WooSunghyeon/paca.

Recent literature has found that an effective method to customize or further improve large language models (LLMs) is to add dynamic adapters, such as low-rank adapters (LoRA) with Mixture-of-Experts (MoE) structures. Though such dynamic adapters incur modest computational complexity, they surprisingly lead to huge inference latency overhead, slowing down the decoding speed by 2.5+ times. In this paper, we analyze the fine-grained costs of the dynamic adapters and find that the fragmented CUDA kernel calls are the root cause. Therefore, we propose LoRA-Switch, a system-algorithm co-designed architecture for efficient dynamic adapters. Unlike most existing dynamic structures that adopt layer-wise or block-wise dynamic routing, LoRA-Switch introduces a token-wise routing mechanism. It switches the LoRA adapters and weights for each token and merges them into the backbone for inference. For efficiency, this switching is implemented with an optimized CUDA kernel, which fuses the merging operations for all LoRA adapters at once. Based on experiments with popular open-source LLMs on common benchmarks, our approach has demonstrated similar accuracy improvement as existing dynamic adapters, while reducing the decoding latency by more than 2.4 times.

The pervasiveness of proprietary language models has raised critical privacy concerns, necessitating advancements in private inference (PI), where computations are performed directly on encrypted data without revealing users' sensitive information. While PI offers a promising solution, its practical deployment is hindered by substantial communication and latency overheads, primarily stemming from nonlinear operations. To address this, we introduce an information-theoretic framework to characterize the role of nonlinearities in decoder-only language models, laying a principled foundation for optimizing transformer-architectures tailored to the demands of PI. By leveraging Shannon's entropy as a quantitative measure, we uncover the previously unexplored dual significance of nonlinearities: beyond ensuring training stability, they are crucial for maintaining attention head diversity. Specifically, we find that their removal triggers two critical failure modes: {\em entropy collapse} in deeper layers that destabilizes training, and {\em entropic overload} in earlier layers that leads to under-utilization of Multi-Head Attention's (MHA) representational capacity. We propose an entropy-guided attention mechanism paired with a novel entropy regularization technique to mitigate entropic overload. Additionally, we explore PI-friendly alternatives to layer normalization for preventing entropy collapse and stabilizing the training of LLMs with reduced-nonlinearities. Our study bridges the gap between information theory and architectural design, establishing entropy dynamics as a principled guide for developing efficient PI architectures. The code and implementation are available at https://github.com/Nandan91/entropy-guided-attention-llm{entropy-guided-llm}.

Large Language Models (LLMs) have demonstrated exceptional performance across diverse domains but are increasingly constrained by high inference latency. Early Exit has emerged as a promising solution to accelerate inference by dynamically bypassing redundant layers. However, in decoder-only architectures, the efficiency of Early Exit is severely bottlenecked by the KV Cache Absence problem, where skipped layers fail to provide the necessary historical states for subsequent tokens. Existing solutions, such as recomputation or masking, either introduce significant latency overhead or incur severe precision loss, failing to bridge the gap between theoretical layer reduction and practical wall-clock speedup. In this paper, we propose River-LLM, a training-free framework that enables seamless token-level Early Exit. River-LLM introduces a lightweight KV-Shared Exit River that allows the backbone's missing KV cache to be naturally generated and preserved during the exit process, eliminating the need for costly recovery operations. Furthermore, we utilize state transition similarity within decoder blocks to predict cumulative KV errors and guide precise exit decisions. Extensive experiments on mathematical reasoning and code generation tasks demonstrate that River-LLM achieves 1.71 to 2.16 times of practical speedup while maintaining high generation quality.

The pervasiveness of proprietary language models has raised privacy concerns for users' sensitive data, emphasizing the need for private inference (PI), where inference is performed directly on encrypted inputs. However, current PI methods face prohibitively higher communication and latency overheads, primarily due to nonlinear operations. In this paper, we present a comprehensive analysis to understand the role of nonlinearities in transformer-based decoder-only language models. We introduce AERO, a four-step architectural optimization framework that refines the existing LLM architecture for efficient PI by systematically removing nonlinearities such as LayerNorm and GELU and reducing FLOPs counts. For the first time, we propose a Softmax-only architecture with significantly fewer FLOPs tailored for efficient PI. Furthermore, we devise a novel entropy regularization technique to improve the performance of Softmax-only models. AERO achieves up to 4.23times communication and 1.94times latency reduction. We validate the effectiveness of AERO by benchmarking it against the state-of-the-art.

BERT based ranking models have achieved superior performance on various information retrieval tasks. However, the large number of parameters and complex self-attention operation come at a significant latency overhead. To remedy this, recent works propose late-interaction architectures, which allow pre-computation of intermediate document representations, thus reducing the runtime latency. Nonetheless, having solved the immediate latency issue, these methods now introduce storage costs and network fetching latency, which limits their adoption in real-life production systems. In this work, we propose the Succinct Document Representation (SDR) scheme that computes highly compressed intermediate document representations, mitigating the storage/network issue. Our approach first reduces the dimension of token representations by encoding them using a novel autoencoder architecture that uses the document's textual content in both the encoding and decoding phases. After this token encoding step, we further reduce the size of entire document representations using a modern quantization technique. Extensive evaluations on passage re-reranking on the MSMARCO dataset show that compared to existing approaches using compressed document representations, our method is highly efficient, achieving 4x-11.6x better compression rates for the same ranking quality.

Huawei's openPangu-Embedded-1B and openPangu-Embedded-7B are variants of the openPangu large language model, designed for efficient deployment on Ascend NPUs. The 7B variant supports three distinct Chain-of-Thought (CoT) reasoning paradigms, namely slow_think, auto_think, and no_think, while the 1B variant operates exclusively in the no_think mode, which employs condensed reasoning for higher efficiency. Although CoT reasoning enhances capability, the generation of extended reasoning traces introduces substantial memory and latency overheads, posing challenges for practical deployment on Ascend NPUs. This paper addresses these computational constraints by leveraging low-bit quantization, which transforms FP16 computations into more efficient integer arithmetic. We introduce a unified low-bit inference framework, supporting INT8 (W8A8) and W4A8 quantization, specifically optimized for openPangu-Embedded models on the Atlas A2. Our comprehensive evaluation on code generation benchmarks (HumanEval and MBPP) demonstrates the efficacy of this approach. INT8 quantization consistently preserves over 90\% of the FP16 baseline accuracy and achieves a 1.5x prefill speedup on the Atlas A2. Furthermore, W4A8 quantization significantly reduces memory consumption, albeit with a moderate trade-off in accuracy. These findings collectively indicate that low-bit quantization effectively facilitates efficient CoT reasoning on Ascend NPUs, maintaining high model fidelity.

Large reasoning models (LRMs) have demonstrated remarkable proficiency in tackling complex reasoning tasks through step-by-step thinking. However, such a lengthy reasoning process incurs substantial computational and latency overheads, hindering the practical deployment of these models. In this work, we present a new perspective on mitigating overthinking in LRMs via black-box adversarial prompting. By treating both open-source LRMs and closed-source APIs as black-box communicators, we investigate how to elicit concise responses without sacrificing accuracy. We introduce AdvPrompt, an iterative refinement framework that generates high-quality adversarial prompts from diverse perspectives. Experiments across multiple benchmarks demonstrate that AdvPrompt consistently reduces token usage while preserving performance. Notably, AdvPrompt achieves a 3x reduction in average response length on simple GSM8K questions for the Qwen3 model series, and delivers an average ~40% token reduction across four benchmarks. For closed-source APIs, AdvPrompt reduces token usage on MATH-500 by 35% for Claude-3.7 and 47% for Gemini-2.5. Further analysis reveals the generalizability of AdvPrompt across various model scales and families, underscoring the potential of black-box prompting as a practical and effective strategy for enhancing LRM efficiency.

The computational sparsity of Mixture-of-Experts (MoE) models enables sub-linear growth in compute cost as model size increases, thus offering a scalable path to training massive neural networks. However, existing implementations suffer from low GPU utilization, significant latency overhead, and a fundamental inability to leverage task locality, primarily due to CPU-managed scheduling, host-initiated communication, and frequent kernel launches. To overcome these limitations, we develop FlashMoE, a fully GPU-resident MoE operator that fuses expert computation and inter-GPU communication into a single persistent GPU kernel. FlashMoE enables fine-grained pipelining of dispatch, compute, and combine phases, eliminating launch overheads and reducing idle gaps. Unlike existing work, FlashMoE eliminates bulk-synchronous collectives for one-sided, device-initiated, inter-GPU (R)DMA transfers, thereby unlocking payload efficiency by eliminating bloated or redundant network payloads in sparsely activated layers. When evaluated on an 8-H100 GPU node with MoE models comprising up to 128 experts and 16K token sequences, FlashMoE achieves up to 9x higher GPU utilization, 6x lower latency, 5.7x higher throughput, and 4x better overlap efficiency compared to state-of-the-art baselines, despite using FP32, whereas the baselines use FP16. FlashMoE shows that principled GPU kernel-hardware co-design is key to unlocking the performance ceiling of large-scale distributed ML. We provide code at https://github.com/osayamenja/FlashMoE.

Large Vision Language Models (LVLMs) have achieved remarkable success in a range of downstream tasks that require multimodal interaction, but their capabilities come with substantial computational and memory overhead, which hinders practical deployment. Among numerous acceleration techniques, post-training quantization is a popular and effective strategy for reducing memory cost and accelerating inference. However, existing LVLM quantization methods typically measure token sensitivity at the modality level, which fails to capture the complex cross-token interactions and falls short in quantitatively measuring the quantization error at the token level. As tokens interact within the model, the distinction between modalities gradually diminishes, suggesting the need for fine-grained calibration. Inspired by axiomatic attribution in mechanistic interpretability, we introduce a fine-grained quantization strategy on Quantization-aware Integrated Gradients (QIG), which leverages integrated gradients to quantitatively evaluate token sensitivity and push the granularity from modality level to token level, reflecting both inter-modality and intra-modality dynamics. Extensive experiments on multiple LVLMs under both W4A8 and W3A16 settings show that our method improves accuracy across models and benchmarks with negligible latency overhead. For example, under 3-bit weight-only quantization, our method improves the average accuracy of LLaVA-onevision-7B by 1.60%, reducing the gap to its full-precision counterpart to only 1.33%. The code is available at https://github.com/ucas-xiang/QIG.

As large language models (LLMs) evolve into autonomous agents, their real-world applicability has expanded significantly, accompanied by new security challenges. Most existing agent defense mechanisms adopt a mandatory checking paradigm, in which security validation is forcibly triggered at predefined stages of the agent lifecycle. In this work, we argue that effective agent security should be intrinsic and selective rather than architecturally decoupled and mandatory. We propose Spider-Sense framework, an event-driven defense framework based on Intrinsic Risk Sensing (IRS), which allows agents to maintain latent vigilance and trigger defenses only upon risk perception. Once triggered, the Spider-Sense invokes a hierarchical defence mechanism that trades off efficiency and precision: it resolves known patterns via lightweight similarity matching while escalating ambiguous cases to deep internal reasoning, thereby eliminating reliance on external models. To facilitate rigorous evaluation, we introduce S^2Bench, a lifecycle-aware benchmark featuring realistic tool execution and multi-stage attacks. Extensive experiments demonstrate that Spider-Sense achieves competitive or superior defense performance, attaining the lowest Attack Success Rate (ASR) and False Positive Rate (FPR), with only a marginal latency overhead of 8.3\%.

Recent advances in Large Language Models (LLMs) have been driven by test-time compute scaling - a strategy that improves reasoning by generating longer, sequential thought processes. While effective, this approach encounters a significant bottleneck as computation increases, where further computation offers only marginal performance gains. We argue this ceiling is not an inherent limit of the model's capability but a flaw in the scaling strategy itself, a phenomenon we term "Tunnel Vision", where a model's imperfect initial steps lock it into a suboptimal reasoning path. To overcome this, we introduce a new scaling paradigm: native thought parallelism. We present ParaThinker, an end-to-end framework that trains an LLM to generate multiple, diverse reasoning paths in parallel and synthesize them into a superior final answer. By exploring different lines of thoughts simultaneously, ParaThinker effectively sidesteps the Tunnel Vision issue and unlocks the model's latent reasoning potential. Our approach demonstrates that scaling compute in parallel (width) is a more effective and efficient way to superior reasoning than simply scaling sequentially (depth). On challenging reasoning benchmarks, ParaThinker achieves substantial accuracy improvements over sequential LLMs (12.3% for 1.5B and 7.5% for 7B models on average with 8 parallel paths), while adding only negligible latency overhead (7.1%). This enables smaller models to surpass much larger counterparts and establishes parallel thinking as a critical, efficient dimension for scaling future LLMs.

Speculative decoding accelerates LLM inference by using a draft model to look ahead, but gains are capped by the cost of autoregressive draft generation: increasing draft size elevates acceptance rates but introduces additional latency overhead exacerbating the speed-accuracy tradeoff. Prior methods (Medusa, Hydra, EAGLE) partially reduce draft cost but either degrade acceptance or introduce overheads that limit scaling. We present Mirror Speculative Decoding (Mirror-SD), an inference algorithm that breaks the latency-acceptance tradeoff. Mirror-SD launches branch-complete rollouts from early-exit signals in parallel with the target model's suffix and explicitly maps computation across heterogeneous accelerators (GPU and NPU) to exploit cross-device parallelism. The draft speculates forward continuations for the target to verify, while the target simultaneously speculates correction paths for the draft, converting speculation into two complementary execution pipelines. To further cut draft latency without weakening acceptance semantics, we add speculative streaming so the draft emits multiple tokens per step. This dual strategy of parallel heterogeneous execution plus multi-token speculative streaming pushes speculative decoding toward its ideal regime of high acceptance with low overhead. On SpecBench with server-scale models from 14B to 66B parameters, Mirror-SD delivers consistent end-to-end gains, achieving 2.8x-5.8x wall-time speedups across diverse tasks and a 30% average relative improvement over the strongest baseline, EAGLE3.

Resolving real-world software engineering (SWE) issues with autonomous agents requires complex, long-horizon reasoning. Current pipelines are bottlenecked by unoptimized demonstration data, sparse execution rewards, and computationally prohibitive inference scaling, which collectively exacerbate token bloat, reward hacking, and policy degradation. We present SWE-TRACE (Trajectory Reduction and Agentic Criteria Evaluation), a unified framework optimizing the SWE agent lifecycle across data curation, reinforcement learning (RL), and test-time inference. First, we introduce an LLM multi-task cascading method, utilizing stepwise oracle verification to distill a 60K-instance Supervised Fine-Tuning (SFT) corpus strictly biased toward token-efficient, shortest-path trajectories. Second, to overcome the instability of sparse outcome rewards, we design a MemoryAugmented Agentic RL pipeline featuring a Rubric-Based Process Reward Model (PRM). An auxiliary Rubric-Agent provides dense, fine-grained heuristic feedback on intermediate steps, guiding the model through long-horizon tasks. Finally, we bridge training and inference by repurposing the PRM for heuristic-guided Test-Time Scaling (TTS). By dynamically evaluating and pruning action candidates at each step, SWE-TRACE achieves superior search efficiency without the latency overhead of standard parallel sampling. Extensive experiments on standard SWE benchmarks demonstrate that SWE-TRACE significantly advances the state-of-the-art, maximizing resolution rates while drastically reducing both token consumption and inference latency.

The Model Context Protocol (MCP) has emerged as a de facto standard for integrating Large Language Models with external tools, yet no formal security analysis of the protocol specification exists. We present the first rigorous security analysis of MCP's architectural design, identifying three fundamental protocol-level vulnerabilities: (1) absence of capability attestation allowing servers to claim arbitrary permissions, (2) bidirectional sampling without origin authentication enabling server-side prompt injection, and (3) implicit trust propagation in multi-server configurations. We implement MCPBench, a novel framework bridging existing agent security benchmarks to MCP-compliant infrastructure, enabling direct measurement of protocol-specific attack surfaces. Through controlled experiments on 847 attack scenarios across five MCP server implementations, we demonstrate that MCP's architectural choices amplify attack success rates by 23--41\% compared to equivalent non-MCP integrations. We propose MCPSec, a backward-compatible protocol extension adding capability attestation and message authentication, reducing attack success rates from 52.8\% to 12.4\% with median latency overhead of 8.3ms per message. Our findings establish that MCP's security weaknesses are architectural rather than implementation-specific, requiring protocol-level remediation.

This paper addresses the growing need for efficient large language models (LLMs) on mobile devices, driven by increasing cloud costs and latency concerns. We focus on designing top-quality LLMs with fewer than a billion parameters, a practical choice for mobile deployment. Contrary to prevailing belief emphasizing the pivotal role of data and parameter quantity in determining model quality, our investigation underscores the significance of model architecture for sub-billion scale LLMs. Leveraging deep and thin architectures, coupled with embedding sharing and grouped-query attention mechanisms, we establish a strong baseline network denoted as MobileLLM, which attains a remarkable 2.7%/4.3% accuracy boost over preceding 125M/350M state-of-the-art models. Additionally, we propose an immediate block-wise weight sharing approach with no increase in model size and only marginal latency overhead. The resultant models, denoted as MobileLLM-LS, demonstrate a further accuracy enhancement of 0.7%/0.8% than MobileLLM 125M/350M. Moreover, MobileLLM model family shows significant improvements compared to previous sub-billion models on chat benchmarks, and demonstrates close correctness to LLaMA-v2 7B in API calling tasks, highlighting the capability of small models for common on-device use cases.

Scaling Large Language Models (LLMs) typically relies on increasing the number of parameters or test-time computations to boost performance. However, these strategies are impractical for edge device deployment due to limited RAM and NPU resources. Despite hardware constraints, deploying performant LLM on edge devices such as smartphone remains crucial for user experience. To address this, we propose MeKi (Memory-based Expert Knowledge Injection), a novel system that scales LLM capacity via storage space rather than FLOPs. MeKi equips each Transformer layer with token-level memory experts that injects pre-stored semantic knowledge into the generation process. To bridge the gap between training capacity and inference efficiency, we employ a re-parameterization strategy to fold parameter matrices used during training into a compact static lookup table. By offloading the knowledge to ROM, MeKi decouples model capacity from computational cost, introducing zero inference latency overhead. Extensive experiments demonstrate that MeKi significantly outperforms dense LLM baselines with identical inference speed, validating the effectiveness of memory-based scaling paradigm for on-device LLMs. Project homepage is at https://github.com/ningding-o/MeKi.

Advanced reasoning typically requires Chain-of-Thought prompting, which is accurate but incurs prohibitive latency and substantial test-time inference costs. The standard alternative, fine-tuning smaller models, often sacrifices interpretability while introducing significant resource and operational overhead. To address these limitations, we introduce Prompt-Level Distillation (PLD). We extract explicit reasoning patterns from a Teacher model and organize them into a structured list of expressive instructions for the Student model's System Prompt. Evaluated using Gemma-3 4B, PLD improved Macro F1 scores on StereoSet (57\% to 90.0\%) and Contract-NLI (67\% to 83\%), while increasing LogiQA accuracy to 70\%. Similar results on Mistral Small 3.1 demonstrate cross-architecture generalizability, enabling these compact models to match frontier performance with negligible latency overhead. These expressive instructions render the decision-making process transparent, allowing for full human verification of logic, making this approach ideal for regulated industries such as law, finance, and content moderation, as well as high-volume use cases and edge devices.

Large Language Models (LLMs) are increasingly deployed in long-context tasks such as reasoning, code generation, and multi-turn dialogue. However, inference over extended contexts is bottlenecked by the Key-Value (KV) cache, whose memory footprint grows linearly with sequence length and dominates latency at each decoding step. While recent KV cache compression methods identify and load important tokens, they focus predominantly on input contexts and fail to address the cumulative attention errors that arise during long decoding. In this paper, we introduce RetroAttention, a novel KV cache update technique that retrospectively revises past attention outputs using newly arrived KV entries from subsequent decoding steps. By maintaining a lightweight output cache, RetroAttention enables past queries to efficiently access more relevant context, while incurring minimal latency overhead. This breaks the fixed-attention-output paradigm and allows continual correction of prior approximations. Extensive experiments on long-generation benchmarks show that RetroAttention consistently outperforms state-of-the-art (SOTA) KV compression methods, increasing effective KV exposure by up to 1.6times and accuracy by up to 21.9\%.

Reasoning large language models (RLLMs) have recently demonstrated remarkable capabilities through structured and multi-step reasoning. While prior research has primarily focused on improving their training and inference strategies, their potential for in-context learning (ICL) remains largely underexplored. To fill this gap, we propose Thinking with Nothinking Calibration (JointThinking), a new ICL paradigm that leverages the structured difference between two reasoning modes, i.e., Thinking and Nothinking, to improve reasoning accuracy. Specifically, our method prompts the model to generate two answers in parallel: one in Thinking mode and the other in Nothinking mode. A second round of Thinking is triggered only when the two initial responses are inconsistent, using a single prompt that incorporates the original question and both candidate answers. Since such disagreement occurs infrequently (e.g., only 6\% in GSM8K), our method performs just one round of reasoning in most cases, resulting in minimal latency overhead. Extensive experiments across multiple reasoning benchmarks demonstrate that JointThinking significantly outperforms few-shot chain-of-thought (CoT) and majority voting with improved answer robustness. Moreover, It achieves comparable in-distribution performance to training-based SOTA method, while substantially outperforming on out-of-distribution tasks. We further conduct a systematic analysis of the calibration mechanism, showing that leveraging different reasoning modes consistently lowers the error rate and highlights the value of structural thinking diversity. Additionally, we observe that the performance gap between actual and ideal reasoning narrows as model size increases in the second round of thinking, indicating the strong scalability of our approach. Finally, we discuss current limitations and outline promising directions for future ICL research in RLLMs.

This paper presents MoE-Infinity, a cost-efficient mixture-of-expert (MoE) serving system that realizes activation-aware expert offloading. MoE-Infinity features sequence-level expert activation tracing, a new approach adept at identifying sparse activations and capturing the temporal locality of MoE inference. By analyzing these traces, MoE-Infinity performs novel activation-aware expert prefetching and caching, substantially reducing the latency overheads usually associated with offloading experts for improved cost performance. Extensive experiments in a cluster show that MoE-Infinity outperforms numerous existing systems and approaches, reducing latency by 4 - 20X and decreasing deployment costs by over 8X for various MoEs. MoE-Infinity's source code is publicly available at https://github.com/TorchMoE/MoE-Infinity

Deformable image registration (DIR) remains a fundamental yet challenging problem in medical image analysis, largely due to the prohibitively high-dimensional deformation space of dense displacement fields and the scarcity of voxel-level supervision. Existing reinforcement learning frameworks often project this space into coarse, low-dimensional representations, limiting their ability to capture spatially variant deformations. We propose MorphSeek, a fine-grained representation-level policy optimization paradigm that reformulates DIR as a spatially continuous optimization process in the latent feature space. MorphSeek introduces a stochastic Gaussian policy head atop the encoder to model a distribution over latent features, facilitating efficient exploration and coarse-to-fine refinement. The framework integrates unsupervised warm-up with weakly supervised fine-tuning through Group Relative Policy Optimization, where multi-trajectory sampling stabilizes training and improves label efficiency. Across three 3D registration benchmarks (OASIS brain MRI, LiTS liver CT, and Abdomen MR-CT), MorphSeek achieves consistent Dice improvements over competitive baselines while maintaining high label efficiency with minimal parameter cost and low step-level latency overhead. Beyond optimizer specifics, MorphSeek advances a representation-level policy learning paradigm that achieves spatially coherent and data-efficient deformation optimization, offering a principled, backbone-agnostic, and optimizer-agnostic solution for scalable visual alignment in high-dimensional settings.

The security of Large Language Model (LLM) applications is fundamentally challenged by "form-first" attacks like prompt injection and jailbreaking, where malicious instructions are embedded within user inputs. Conventional defenses, which rely on post hoc output filtering, are often brittle and fail to address the root cause: the model's inability to distinguish trusted instructions from untrusted data. This paper proposes Countermind, a multi-layered security architecture intended to shift defenses from a reactive, post hoc posture to a proactive, pre-inference, and intra-inference enforcement model. The architecture proposes a fortified perimeter designed to structurally validate and transform all inputs, and an internal governance mechanism intended to constrain the model's semantic processing pathways before an output is generated. The primary contributions of this work are conceptual designs for: (1) A Semantic Boundary Logic (SBL) with a mandatory, time-coupled Text Crypter intended to reduce the plaintext prompt injection attack surface, provided all ingestion paths are enforced. (2) A Parameter-Space Restriction (PSR) mechanism, leveraging principles from representation engineering, to dynamically control the LLM's access to internal semantic clusters, with the goal of mitigating semantic drift and dangerous emergent behaviors. (3) A Secure, Self-Regulating Core that uses an OODA loop and a learning security module to adapt its defenses based on an immutable audit log. (4) A Multimodal Input Sandbox and Context-Defense mechanisms to address threats from non-textual data and long-term semantic poisoning. This paper outlines an evaluation plan designed to quantify the proposed architecture's effectiveness in reducing the Attack Success Rate (ASR) for form-first attacks and to measure its potential latency overhead.

Browser-using agents (BUAs) are an emerging class of AI agents that interact with web browsers in human-like ways, including clicking, scrolling, filling forms, and navigating across pages. While these agents help automate repetitive online tasks, they are vulnerable to prompt injection attacks that trick an agent into performing undesired actions, such as leaking private information or issuing unintended state-changing requests. We propose ceLLMate, a browser-level sandboxing framework that restricts the agent's ambient authority and reduces the blast radius of prompt injections. We address the semantic gap challenge that is fundamental to BUAs -- writing and enforcing security policies for low-level UI tools like clicks and keystrokes is brittle and error-prone. Our core insight is to perform sandboxing at the HTTP layer because all side-effecting UI operations will result in network communication to the website's backend. We implement ceLLMate as an agent-agnostic browser extension and demonstrate how it enables sandboxing policies that block prompt injection attacks in the WASP benchmark with 7.25--15% latency overhead.

Lifelong user interest modeling is crucial for industrial recommender systems, yet existing approaches rely predominantly on ID-based features, suffering from poor generalization on long-tail items and limited semantic expressiveness. While recent work explores multimodal representations for behavior retrieval in the General Search Unit (GSU), they often neglect multimodal integration in the fine-grained modeling stage -- the Exact Search Unit (ESU). In this work, we present a systematic analysis of how to effectively leverage multimodal signals across both stages of the two-stage lifelong modeling framework. Our key insight is that simplicity suffices in the GSU: lightweight cosine similarity with high-quality multimodal embeddings outperforms complex retrieval mechanisms. In contrast, the ESU demands richer multimodal sequence modeling and effective ID-multimodal fusion to unlock its full potential. Guided by these principles, we propose MUSE, a simple yet effective multimodal search-based framework. MUSE has been deployed in Taobao display advertising system, enabling 100K-length user behavior sequence modeling and delivering significant gains in top-line metrics with negligible online latency overhead. To foster community research, we share industrial deployment practices and open-source the first large-scale dataset featuring ultra-long behavior sequences paired with high-quality multimodal embeddings. Our code and data is available at https://taobao-mm.github.io.

Recent advancements in function calling and tool use have significantly enhanced the capabilities of large language models (LLMs) by enabling them to interact with external information sources and execute complex tasks. However, the limited context window of LLMs presents challenges when a large number of tools are available, necessitating efficient methods to manage prompt length and maintain accuracy. Existing approaches, such as fine-tuning LLMs or leveraging their reasoning capabilities, either require frequent retraining or incur significant latency overhead. A more efficient solution involves training smaller models to retrieve the most relevant tools for a given query, although this requires high quality, domain-specific data. To address those challenges, we present a novel framework for generating synthetic data for tool retrieval applications and an efficient data-driven tool retrieval strategy using small encoder models. Empowered by LLMs, we create ToolBank, a new tool retrieval dataset that reflects real human user usages. For tool retrieval methodologies, we propose novel approaches: (1) Tool2Vec: usage-driven tool embedding generation for tool retrieval, (2) ToolRefiner: a staged retrieval method that iteratively improves the quality of retrieved tools, and (3) MLC: framing tool retrieval as a multi-label classification problem. With these new methods, we achieve improvements of up to 27.28 in Recall@K on the ToolBench dataset and 30.5 in Recall@K on ToolBank. Additionally, we present further experimental results to rigorously validate our methods. Our code is available at https://github.com/SqueezeAILab/Tool2Vec

The demand for deploying large language models(LLMs) on mobile devices continues to increase, driven by escalating data security concerns and cloud costs. However, network bandwidth and memory limitations pose challenges for deploying billion-level models on mobile devices. In this study, we investigate the outputs of different layers across various scales of LLMs and found that the outputs of most layers exhibit significant similarity. Moreover, this similarity becomes more pronounced as the model size increases, indicating substantial redundancy in the depth direction of the LLMs. Based on this observation, we propose an efficient model volume compression strategy, termed FoldGPT, which combines block removal and block parameter sharing.This strategy consists of three parts: (1) Based on the learnable gating parameters, we determine the block importance ranking while modeling the coupling effect between blocks. Then we delete some redundant layers based on the given removal rate. (2) For the retained blocks, we apply a specially designed group parameter sharing strategy, where blocks within the same group share identical weights, significantly compressing the number of parameters and slightly reducing latency overhead. (3) After sharing these Blocks, we "cure" the mismatch caused by sparsity with a minor amount of fine-tuning and introduce a tail-layer distillation strategy to improve the performance. Experiments demonstrate that FoldGPT outperforms previous state-of-the-art(SOTA) methods in efficient model compression, demonstrating the feasibility of achieving model lightweighting through straightforward block removal and parameter sharing.

This paper introduces JaxPruner, an open-source JAX-based pruning and sparse training library for machine learning research. JaxPruner aims to accelerate research on sparse neural networks by providing concise implementations of popular pruning and sparse training algorithms with minimal memory and latency overhead. Algorithms implemented in JaxPruner use a common API and work seamlessly with the popular optimization library Optax, which, in turn, enables easy integration with existing JAX based libraries. We demonstrate this ease of integration by providing examples in four different codebases: Scenic, t5x, Dopamine and FedJAX and provide baseline experiments on popular benchmarks.

Secure multi-party computation (MPC) enables computation directly on encrypted data and protects both data and model privacy in deep learning inference. However, existing neural network architectures, including Vision Transformers (ViTs), are not designed or optimized for MPC and incur significant latency overhead. We observe Softmax accounts for the major latency bottleneck due to a high communication complexity, but can be selectively replaced or linearized without compromising the model accuracy. Hence, in this paper, we propose an MPC-friendly ViT, dubbed MPCViT, to enable accurate yet efficient ViT inference in MPC. Based on a systematic latency and accuracy evaluation of the Softmax attention and other attention variants, we propose a heterogeneous attention optimization space. We also develop a simple yet effective MPC-aware neural architecture search algorithm for fast Pareto optimization. To further boost the inference efficiency, we propose MPCViT+, to jointly optimize the Softmax attention and other network components, including GeLU, matrix multiplication, etc. With extensive experiments, we demonstrate that MPCViT achieves 1.9%, 1.3% and 3.6% higher accuracy with 6.2x, 2.9x and 1.9x latency reduction compared with baseline ViT, MPCFormer and THE-X on the Tiny-ImageNet dataset, respectively. MPCViT+ further achieves a better Pareto front compared with MPCViT. The code and models for evaluation are available at https://github.com/PKU-SEC-Lab/mpcvit.

Despite rapid progress in autoregressive video diffusion, an emerging system algorithm bottleneck limits both deployability and generation capability: KV cache memory. In autoregressive video generation models, the KV cache grows with generation history and quickly dominates GPU memory, often exceeding 30 GB, preventing deployment on widely available hardware. More critically, constrained KV cache budgets restrict the effective working memory, directly degrading long horizon consistency in identity, layout, and motion. To address this challenge, we present Quant VideoGen (QVG), a training free KV cache quantization framework for autoregressive video diffusion models. QVG leverages video spatiotemporal redundancy through Semantic Aware Smoothing, producing low magnitude, quantization friendly residuals. It further introduces Progressive Residual Quantization, a coarse to fine multi stage scheme that reduces quantization error while enabling a smooth quality memory trade off. Across LongCat Video, HY WorldPlay, and Self Forcing benchmarks, QVG establishes a new Pareto frontier between quality and memory efficiency, reducing KV cache memory by up to 7.0 times with less than 4% end to end latency overhead while consistently outperforming existing baselines in generation quality.

Collocating deep learning training tasks improves GPU utilization but causes drastic slowdowns due to resource contention and risks Out-of-Memory (OOM) failures. Accurate memory estimation is essential for robust collocation, while GPU utilization -- a key proxy for resource contention -- enables interference-aware scheduling to reduce slowdowns and improve throughput. Existing GPU memory estimators span three paradigms -- analytical models, CPU-side libraries, and ML-based estimators -- each with distinct limitations: dependence on detailed model specifications, intrusive integration, poor generalization, and varying latency overhead. GPU heterogeneity further complicates estimation, as identical tasks can exhibit markedly different memory footprints across hardware generations. GPU utilization remains comparatively understudied, further complicated by the non-additive nature of utilization metrics and hardware sensitivity. We conduct a systematic analysis of representative estimators from each paradigm -- Horus, PyTorch FakeTensor, and our lightweight ML-based estimator -- evaluating accuracy, generalizability, and practical overhead. We construct a synthetic dataset spanning MLPs, CNNs, and Transformers with controlled architectural variations, and train MLP- and Transformer-based estimators for memory prediction. We further experiment with utilization estimation on the same dataset. Our evaluation reveals key tradeoffs and validates estimators against real-world unseen models. Significant challenges remain: analytical models are hardware-dependent, CPU-side libraries impose intrusive integration costs, and ML-based estimators struggle with cross-architecture generalization. We release all datasets, tools, and artifacts to support further research.

Real-time video restoration (VR) for live streams requires high-resolution outputs under strict per-frame latency constraints. Existing one-step diffusion-based VR models remain difficult to deploy on consumer-grade GPUs due to two main bottlenecks: quadratic spatial attention at high resolutions and the latency-memory overhead of large video autoencoders. We present SwiftVR, a streaming one-step generative VR framework that reduces both bottlenecks under a causal chunk-wise protocol. For attention, mask-free shifted-window self-attention gathers each spatial window into a dense tensor via deterministic indexing, keeping all attention calls on the dense scaled dot-product attention path without masks, cyclic shifts, padding, or hardware-specific sparse kernels. Because SwiftVR uses only standard dense SDPA calls, the trained model transfers to consumer GPUs without retraining or custom kernels. For autoencoding, a lightweight Restoration-aware Autoencoder enables fast chunk-wise decoding while preserving reconstruction quality. On a single H100, SwiftVR sustains 31~FPS at 2560x1440 and 14~FPS at 3840x2160, whereas all compared diffusion-based VR baselines exceed the memory limit at 4K. On a consumer RTX~5090, SwiftVR reaches 26~FPS at 1920x1080. To our knowledge, SwiftVR is the first generative VR model to achieve real-time 1080p streaming on a consumer-grade GPU, while attaining strong no-reference perceptual quality with lower inference cost. Project is available at https://h-oliday.github.io/SwiftVR.

Integrating large language models (LLMs) into automatic speech recognition (ASR) has become a dominant paradigm. Although recent LLM-based ASR models have shown promising performance on public benchmarks, it remains challenging to balance recognition quality with latency and overhead, while hallucinations further limit real-world deployment. In this study, we revisit LLM-based ASR from an entropy allocation perspective and introduce three metrics to characterize how training paradigms allocate entropy reduction between the speech encoder and the LLM. To remedy entropy-allocation inefficiencies in prevailing approaches, we propose a principled multi-stage training strategy grounded in capability-boundary awareness, optimizing parameter efficiency and hallucination robustness. Specifically, we redesign the pretraining strategy to alleviate the speech-text modality gap, and further introduce an iterative asynchronous SFT stage between alignment and joint SFT to preserve functional decoupling and constrain encoder representation drift. Experiments on Mandarin and English benchmarks show that our method achieves competitive performance with state-of-the-art models using only 2.3B parameters, while also effectively mitigating hallucinations through our decoupling-oriented design.

Efficient scheduling of LLM inference tasks is essential for achieving low latency and high throughput, particularly with the growing use of reasoning-capable LLMs. Traditional strategies like First-Come-First-Serve (FCFS) often suffer from Head-of-Line (HOL) blocking, where long-running tasks delay shorter ones queued behind them. In this paper, we introduce PARS, a prompt-aware LLM task scheduler that improves serving efficiency by approximating shortest-job-first (SJF) scheduling through pairwise ranking with margin ranking loss. PARS focuses on impactful scheduling decisions and is seamlessly integrated into the state-of-the-art LLM serving system vLLM. It effectively predicts response-length-based task ordering, reducing latency with minimal overhead. Extensive experiments across multiple LLMs and real-world inference datasets show that PARS significantly improves performance, including for reasoning workloads. Furthermore, our cross-model evaluations demonstrate that the design generalizes well, enabling effective scheduling even when predictors are trained on different LLMs.

LLM-based agents solve complex tasks through iterative reasoning, tool use, and environment interaction, where each intermediate thought directly shapes subsequent actions. Small deviations in these thoughts can therefore propagate into unsafe behaviors, yet existing guardrails typically operate only on final outputs or require intrusive model modifications. We introduce Thought-Aligner, a lightweight plug-in safety model that performs causal correction on unsafe thoughts before action execution, without altering the underlying agent. The corrected thoughts are fed back into the agent, steering its decision process and tool use toward safer trajectories. Because it operates solely at the thought level, Thought-Aligner is model-agnostic and can be integrated into diverse agent frameworks. We train Thought-Aligner via two-stage contrastive learning on paired safe and unsafe thoughts generated across ten risk scenarios. Experiments on diverse agent-safety benchmarks and six LLMs show that Thought-Aligner increases behavioral safety from about 50% without protection to around 90% on average, exceeding state-of-the-art guardrails by roughly 23%, while also improving helpfulness by about 5%. The method incurs low per-step latency and minimal overhead, enabling scalable and practical deployment. We publicly release Thought-Aligner-7B at https://huggingface.co/WhitzardAgent/Thought-Aligner-7B.

The Key-Value (KV) cache in generative large language models (LLMs) introduces substantial memory overhead. Existing works mitigate this burden by offloading or compressing the KV cache. However, loading the entire cache incurs significant latency due to PCIe bandwidth bottlenecks in CPU-GPU communication, while aggressive compression causes notable performance degradation. We identify that certain layers in the LLM need to maintain global information and are unsuitable for selective loading. In contrast, other layers primarily focus on a few tokens with dominant activations that potentially incur substantial quantization error. This observation leads to a key insight that loading dominant tokens and quantizing all tokens can complement each other. Building on this insight, we propose a hybrid compression method, TailorKV, which seamlessly integrates quantization and offloading. TailorKV develops an inference framework along with a hardware-friendly implementation that leverages these complementary characteristics. Extensive long-context evaluations exhibit that TailorKV achieves nearly lossless performance under aggressive compression settings, outperforming the state-of-the-art. Particularly, the Llama-3.1-8B with 128k context can be served within a single RTX 3090 GPU, reaching 82 ms per token during decoding.

Large reasoning models (LRMs) are proficient at generating explicit, step-by-step reasoning sequences before producing final answers. However, such detailed reasoning can introduce substantial computational overhead and latency, particularly for simple problems. To address this over-thinking problem, we explore how to equip LRMs with adaptive thinking capabilities: enabling them to dynamically decide whether or not to engage in explicit reasoning based on problem complexity. Building on R1-style distilled models, we observe that inserting a simple ellipsis ("...") into the prompt can stochastically trigger either a thinking or no-thinking mode, revealing a latent controllability in the reasoning behavior. Leveraging this property, we propose AutoThink, a multi-stage reinforcement learning (RL) framework that progressively optimizes reasoning policies via stage-wise reward shaping. AutoThink learns to invoke explicit reasoning only when necessary, while defaulting to succinct responses for simpler tasks. Experiments on five mainstream mathematical benchmarks demonstrate that AutoThink achieves favorable accuracy-efficiency trade-offs compared to recent prompting and RL-based pruning methods. It can be seamlessly integrated into any R1-style model, including both distilled and further fine-tuned variants. Notably, AutoThink improves relative accuracy by 6.4 percent while reducing token usage by 52 percent on DeepSeek-R1-Distill-Qwen-1.5B, establishing a scalable and adaptive reasoning paradigm for LRMs. Project Page: https://github.com/ScienceOne-AI/AutoThink.

As Large Language Models (LLMs) are increasingly deployed as autonomous agents, they face a critical scalability bottleneck known as the "Generalization-Specialization Dilemma." Monolithic agents equipped with extensive toolkits suffer from context pollution and attention decay, leading to hallucinations. Conversely, static multi-agent swarms introduce significant latency and resource overhead. This paper introduces a Self-Evolving Concierge System, a novel architecture utilizing a Dynamic Mixture of Experts (DMoE) approach. Unlike recent self-improving agents that rewrite their own codebase, our system preserves stability by dynamically restructuring its runtime environment: "hiring" specialized sub-agents based on real-time conversation analysis. We introduce an asynchronous "Meta-Cognition Engine" that detects capability gaps, a Least Recently Used (LRU) eviction policy for resource constraints, and a novel "Surgical History Pruning" mechanism to mitigate refusal bias. Experimental results demonstrate that this architecture maintains high task success rates while minimizing token consumption compared to static agent swarms.

Attention steering is an important technique for controlling model focus, enabling capabilities such as prompt highlighting, where the model prioritises user-specified text. However, existing attention steering methods require explicit storage of the full attention matrix, making them incompatible with memory-efficient implementations like FlashAttention. We introduce Spectral Editing Key Amplification (SEKA), a training-free steering method that tackles this by directly editing key embeddings before attention computation. SEKA uses spectral decomposition to steer key embeddings towards latent directions that amplify attention scores for certain tokens. We extend this to Adaptive SEKA (AdaSEKA), a query-adaptive variant that uses a training-free routing mechanism to dynamically combine multiple expert subspaces based on the prompt's semantic intent. Our experiments show both methods significantly outperform strong baselines on standard steering benchmarks while adding much lower latency and memory overhead, in compatibility with optimised attention.

Maintaining the safety of large language models (LLMs) is crucial as they are increasingly deployed in real-world applications. Existing safety guardrails typically rely on single-pass classification or, more recently, distilled reasoning. Reasoning-based guardrails significantly outperform classification-only baselines, but they incur substantial query latency and token overhead that make them impractical for highthroughput deployment. To address this challenge, we propose COLAGUARD, a guardrail model that transfers multi-step safety reasoning into a continuous latent space through a stage-wise training curriculum, enabling direct hidden-state propagation at inference. Evaluated on ten prompt- and response-moderation settings spanning eight safety benchmarks, COLAGUARD improves macro-F1 by 8.24 points over Llama Guard 3 and matches our explicit reasoning baseline, GuardReasoner, in macroF1 while delivering a 12.9X speedup and 22.4X reduction in token usage. Our results suggest that latent reasoning offers a practical alternative to explicit rationale generation for deployable guardrails, jointly improving safety robustness and inference efficiency rather than treating them as competing objectives.

Quantization techniques such as BitsAndBytes, AWQ, and GPTQ are widely used as a standard method in deploying large language models but often degrades accuracy when using low-bit representations, e.g., 4 bits. Low-rank correction methods (e.g., LQER, QERA, ASER) has been proposed to mitigate this issue, however, they restore all layers and insert error-correction modules into every decoder block, which increases latency and memory overhead. To address this limitation, we propose GlowQ, a group-shared low-rank approximation for quantized LLMs that caches a single shared right factor per input-sharing group and restores only the groups or layers that yield the highest accuracy benefit. GlowQ computes the high-precision projection once per input-sharing group and reuses it across its modules, reducing parameter and memory overhead, and retaining the expressivity of layer-specific corrections. We also propose a selective variant, GlowQ-S, that applies the cached shared module only where it provides the largest benefit. Compared with strong baselines, our approach reduces TTFB by (5.6%) and increases throughput by (9.6%) on average, while reducing perplexity on WikiText-2 by (0.17%) and increasing downstream accuracy by 0.42 percentage points. The selective model GlowQ-S further reduces latency, cutting TTFB by (23.4%) and increasing throughput by (37.4%), while maintaining accuracy within 0.2 percentage points on average.

Repository-level code comprehension and knowledge sharing remain core challenges in software engineering. Large language models (LLMs) have shown promise by generating explanations of program structure and logic. However, these approaches still face limitations: First, relevant knowledge is distributed across multiple files within a repository, aka semantic fragmentation. Second, retrieval inefficiency and attention saturation degrade performance in RAG pipelines, where long, unaligned contexts overwhelm attention. Third, repository specific training data is scarce and often outdated. Finally, proprietary LLMs hinder industrial adoption due to privacy and deployment constraints. To address these issues, we propose Key-Augmented Neural Triggers (KANT), a novel approach that embeds knowledge anchors into both training and inference. Unlike prior methods, KANT enables internal access to repository specific knowledge, reducing fragmentation and grounding inference in localized context. Moreover, we synthesize specialized data directly from code. At inference, knowledge anchors replace verbose context, reducing token overhead and latency while supporting efficient, on premise deployment. We evaluate KANT via: a qualitative human evaluation of the synthesized dataset's intent coverage and quality across five dimensions; compare against SOTA baselines across five qualitative dimensions and inference speed; and replication across different LLMs to assess generalizability. Results show that the synthetic training data aligned with information-seeking needs. KANT achieved over 60% preference from human annotators and a LocalStack expert (preferring 79% of cases). Also, KANT reduced inference latency by up to 85% across all models. Overall, it is well-suited for scalable, low-latency, on-premise deployments, providing a strong foundation for code comprehension.

Recent studies show that the visual place recognition (VPR) method using pre-trained visual foundation models can achieve promising performance. In our previous work, we propose a novel method to realize seamless adaptation of foundation models to VPR (SelaVPR). This method can produce both global and local features that focus on discriminative landmarks to recognize places for two-stage VPR by a parameter-efficient adaptation approach. Although SelaVPR has achieved competitive results, we argue that the previous adaptation is inefficient in training time and GPU memory usage, and the re-ranking paradigm is also costly in retrieval latency and storage usage. In pursuit of higher efficiency and better performance, we propose an extension of the SelaVPR, called SelaVPR++. Concretely, we first design a parameter-, time-, and memory-efficient adaptation method that uses lightweight multi-scale convolution (MultiConv) adapters to refine intermediate features from the frozen foundation backbone. This adaptation method does not back-propagate gradients through the backbone during training, and the MultiConv adapter facilitates feature interactions along the spatial axes and introduces proper local priors, thus achieving higher efficiency and better performance. Moreover, we propose an innovative re-ranking paradigm for more efficient VPR. Instead of relying on local features for re-ranking, which incurs huge overhead in latency and storage, we employ compact binary features for initial retrieval and robust floating-point (global) features for re-ranking. To obtain such binary features, we propose a similarity-constrained deep hashing method, which can be easily integrated into the VPR pipeline. Finally, we improve our training strategy and unify the training protocol of several common training datasets to merge them for better training of VPR models. Extensive experiments show that ......

Streaming Video Large Language Models (VideoLLMs) have demonstrated impressive performance across various video understanding tasks, but they face significant challenges in real-time deployment due to the high computational cost of processing dense visual tokens from continuous video streams. In streaming video scenarios, the primary bottleneck lies in the Vision Transformer (ViT) encoding stage, where redundant processing of temporally similar frames leads to inefficiency. Additionally, inflated token sequences during LLM pre-filling further exacerbate latency and memory overhead. To address these challenges, we propose Streaming Token Compression (STC), a plug-and-play hierarchical framework that seamlessly integrates into existing streaming VideoLLMs, optimizing both ViT encoding and LLM pre-filling stages to accelerate processing. STC introduces two token-level accelerators: STC-Cacher, which reduces ViT encoding overhead by caching and reusing features from temporally similar frames, and STC-Pruner, which compresses the visual token sequence before it enters the LLM, preserving only the most salient tokens based on both spatial and temporal relevance. Extensive experiments on four baseline streaming VideoLLMs across five benchmarks demonstrate that STC outperforms other compression methods. Notably, STC retains up to 99\% of accuracy on the ReKV framework while reducing ViT encoding latency and LLM pre-filling latency by 24.5\% and 45.3\%.

Vision-based autonomous navigation systems rely on fast and accurate object detection algorithms to avoid obstacles. Algorithms and sensors designed for such systems need to be computationally efficient, due to the limited energy of the hardware used for deployment. Biologically inspired event cameras are a good candidate as a vision sensor for such systems due to their speed, energy efficiency, and robustness to varying lighting conditions. However, traditional computer vision algorithms fail to work on event-based outputs, as they lack photometric features such as light intensity and texture. In this work, we propose a novel technique that utilizes the temporal information inherently present in the events to efficiently detect moving objects. Our technique consists of a lightweight spiking neural architecture that is able to separate events based on the speed of the corresponding objects. These separated events are then further grouped spatially to determine object boundaries. This method of object detection is both asynchronous and robust to camera noise. In addition, it shows good performance in scenarios with events generated by static objects in the background, where existing event-based algorithms fail. We show that by utilizing our architecture, autonomous navigation systems can have minimal latency and energy overheads for performing object detection.

Agentic memory systems enable large language model (LLM) agents to maintain state across long interactions, supporting long-horizon reasoning and personalization beyond fixed context windows. Despite rapid architectural development, the empirical foundations of these systems remain fragile: existing benchmarks are often underscaled, evaluation metrics are misaligned with semantic utility, performance varies significantly across backbone models, and system-level costs are frequently overlooked. This survey presents a structured analysis of agentic memory from both architectural and system perspectives. We first introduce a concise taxonomy of MAG systems based on four memory structures. Then, we analyze key pain points limiting current systems, including benchmark saturation effects, metric validity and judge sensitivity, backbone-dependent accuracy, and the latency and throughput overhead introduced by memory maintenance. By connecting the memory structure to empirical limitations, this survey clarifies why current agentic memory systems often underperform their theoretical promise and outlines directions for more reliable evaluation and scalable system design.

Limited by the context window size of Large Language Models(LLMs), handling various tasks with input tokens exceeding the upper limit has been challenging, whether it is a simple direct retrieval task or a complex multi-hop reasoning task. Although various methods have been proposed to enhance the long-context processing capabilities of LLMs, they either incur substantial post-training costs, or require additional tool modules(e.g.,RAG), or have not shown significant improvement in realistic tasks. Our work observes the correlation between the attention distribution and generated answers across each layer, and establishes the attention allocation aligns with retrieval-augmented capabilities through experiments. Drawing on the above insights, we propose a novel method InfiniRetri that leverages the LLMs's own attention information to enable accurate retrieval across inputs of infinitely length. Our evaluations indicate that InfiniRetri achieves 100% accuracy in the Needle-In-a-Haystack(NIH) test over 1M tokens using a 0.5B parameter model, surpassing other method or larger models and setting a new state-of-the-art(SOTA). Moreover, our method achieves significant performance improvements on real-world benchmarks, with a maximum 288% improvement. In addition, InfiniRetri can be applied to any Transformer-based LLMs without additional training and substantially reduces inference latency and compute overhead in long texts. In summary, our comprehensive studies show InfiniRetri's potential for practical applications and creates a paradigm for retrievaling information using LLMs own capabilities under infinite-length tokens. Code will be released in link.

Recent advances in large language models (LLMs) have predominantly focused on maximizing accuracy and reasoning capabilities, often overlooking crucial computational efficiency considerations. While this approach has yielded impressive accuracy improvements, it has led to methods that may be impractical for real-world deployment due to computational overhead and latency constraints. This paper investigates the potential synergy between reasoning enhancement and computational efficiency by analyzing the integration of two contrasting approaches: Quiet-STaR (Self-Taught Reasoner) and REBASE (REward BAlanced SEarch). Through comprehensive empirical analysis using the Mistral-7B model on the GSM8K dataset, we demonstrate that while each method excels in its primary objective-Quiet-STaR achieving superior accuracy (32.03%) despite high computational cost (554.66s runtime, 12.73T FLOPs), and REBASE providing exceptional efficiency (8.47s runtime, 2.35T FLOPs) while maintaining baseline-comparable accuracy (10.94%)-their integration reveals fundamental challenges in reconciling reasoning depth with computational efficiency. The combined approach unexpectedly results in degraded performance (9.38% accuracy, 143.66s runtime), highlighting critical insights about the complex interplay between reasoning enhancement and efficiency optimization in LLMs. Our findings illuminate the need for novel architectures and algorithms specifically designed to bridge the gap between these competing objectives, while providing concrete directions for future research in compute-efficient reasoning methods.

Serving many task-specialized LLM variants is often limited by the large size of fine-tuned checkpoints and the resulting cold-start latency. Since fine-tuned weights differ from their base model by relatively small structured residuals, a natural approach is to represent them as compressed deltas. We propose a simple 1-bit delta scheme that stores only the sign of the weight difference together with lightweight per-axis (row/column) FP16 scaling factors, learned from a small calibration set. This design preserves the compactness of 1-bit deltas while more accurately capturing variation across weight dimensions, leading to improved reconstruction quality over scalar alternatives. From a systems perspective, a streamlined loader that transfers packed deltas in a single operation per module reduces cold-start latency and storage overhead, with artifacts several times smaller than a full FP16 checkpoint. The method is drop-in, requires minimal calibration data, and maintains inference efficiency by avoiding dense reconstruction. Our experimental setup and source code are available at https://github.com/kuiumdjiev/Per-Axis-Weight-Deltas-for-Frequent-Model-Updates.

Vision-language-action (VLA) models have significantly advanced robotic manipulation by integrating vision-language models (VLMs), and action decoders into a unified architecture. However, their deployment on resource-constrained edge devices, such as mobile robots or embedded systems (e.g., Jetson Orin Nano), remains challenging due to high computational demands, especially in real-world scenarios where power, latency, and computational resources are critical. To close this gap, we introduce Nano-scale Vision-Language Action (NanoVLA), a family of lightweight VLA architectures that achieve high performance with minimal resources. Our core innovations include: (1) vision-language decoupling that moves conventional early vision and language inputs fusion in VLM to late stage, achieving better performance while enabling caching and reduce inference overhead and latency; (2) long-short action chunking to ensure smooth, coherent multi-step planning without sacrificing real-time responsiveness; (3) dynamic routing that adaptively assigns lightweight or heavy backbones based on task complexity, further optimizing inference efficiency. Experimental results on several benchmarks, as well as real-world deployments, demonstrate that NanoVLA achieves up to 52x faster inference on edge devices compared to previous state-of-the-art VLA models, with 98% less parameters while maintaining or surpassing their task accuracy and generalization. Ablation studies confirm that our decoupling strategy preserves cross-task transferability, and the routing module enhances cost-performance trade-offs, enabling practical, high-precision robotic manipulation on resource-constrained hardware.

Video Temporal Grounding (TG) aims to temporally locate video segments matching a natural language description (a query) in a long video. While Vision-Language Models (VLMs) are effective at holistic semantic matching, they often struggle with fine-grained temporal localisation. Recently, Group Relative Policy Optimisation (GRPO) reformulates the inference process as a reinforcement learning task, enabling fine-grained grounding and achieving strong in-domain performance. However, GRPO relies on labelled data, making it unsuitable in unlabelled domains. Moreover, because videos are large and expensive to store and process, performing full-scale adaptation introduces prohibitive latency and computational overhead, making it impractical for real-time deployment. To overcome both problems, we introduce a Data-Efficient Unlabelled Cross-domain Temporal Grounding method, from which a model is first trained on a labelled source domain, then adapted to a target domain using only a small number of unlabelled videos from the target domain. This approach eliminates the need for target annotation and keeps both computational and storage overhead low enough to run in real time. Specifically, we introduce. Uncertainty-quantified Rollout Policy Adaptation (URPA) for cross-domain knowledge transfer in learning video temporal grounding without target labels. URPA generates multiple candidate predictions using GRPO rollouts, averages them to form a pseudo label, and estimates confidence from the variance across these rollouts. This confidence then weights the training rewards, guiding the model to focus on reliable supervision. Experiments on three datasets across six cross-domain settings show that URPA generalises well using only a few unlabelled target videos. Codes will be released once published.

Code generation using large language models (LLMs) is highly sensitive to the choice of tokens during decoding, especially at points of uncertainty that critically affect the generated program's logic. Conventional decoding methods such as greedy search and beam search apply uniform treatment to all tokens, neglecting the unique uncertainty characteristics inherent in code generation, which can result in suboptimal outputs. In this work, we conduct an empirical analysis demonstrating that a significant portion of generation errors arises from incorrect token ranking at high-uncertainty steps, where the ground truth token exists in the candidate set but fails to be ranked first. Inspired by this insight, we introduce AdaDec, an adaptive decoding framework guided by token-level uncertainty quantified via Shannon entropy. AdaDec dynamically learns uncertainty thresholds tailored to each model and employs a pause-then-rerank mechanism with lookahead when the uncertainty surpasses these thresholds. Evaluation on the HumanEval and MBPP benchmarks reveals that AdaDec achieves up to a 15.5% improvement in Pass@1 accuracy compared to greedy decoding, matches or outperforms traditional beam search, and reduces both computational overhead and latency through targeted, selective pausing. Our findings suggest that uncertainty-aware adaptive decoding holds considerable potential for enhancing both the reliability and efficiency of code generation with LLMs.

Test-time scaling improves large language model performance by adding extra compute during decoding. Best-of-N (BoN) sampling serves as a common scaling technique, broadening the search space for finding better solutions from the model distribution. However, traditional BoN requires N full generations, leading to high GPU memory overhead and time latency. Moreover, some methods depend on reward models, adding computational cost and limiting domain generalization. In this paper, we propose Self-Truncation Best-of-N (ST-BoN), a novel decoding method that avoids fully generating all samplings and eliminates the need for reward models. ST-BoN introduces early sampling consistency to estimate the most promising sample, truncating suboptimal ones to free memory and accelerate inference. This pushes the sampling-efficient test-time scaling. Compared to traditional BoN, ST-BoN can reduce dynamic GPU memory overhead by over 90% and time latency by 50%, while achieving comparable or even better performance across reasoning and open-ended domains.

Web proxies such as NGINX commonly rely on least-recently-used (LRU) eviction, which is size agnostic and can thrash under periodic bursts and mixed object sizes. We introduce Cold-RL, a learned eviction policy for NGINX that replaces LRU's forced-expire path with a dueling Deep Q-Network served by an ONNX sidecar within a strict microsecond budget. On each eviction, Cold-RL samples the K least-recently-used objects, extracts six lightweight features (age, size, hit count, inter-arrival time, remaining TTL, and last origin RTT), and requests a bitmask of victims; a hard timeout of 500 microseconds triggers immediate fallback to native LRU. Policies are trained offline by replaying NGINX access logs through a cache simulator with a simple reward: a retained object earns one point if it is hit again before TTL expiry. We compare against LRU, LFU, size-based, adaptive LRU, and a hybrid baseline on two adversarial workloads. With a 25 MB cache, Cold-RL raises hit ratio from 0.1436 to 0.3538, a 146 percent improvement over the best classical baseline; at 100 MB, from 0.7530 to 0.8675, a 15 percent gain; and at 400 MB it matches classical methods (about 0.918). Inference adds less than 2 percent CPU overhead and keeps 95th percentile eviction latency within budget. To our knowledge, this is the first reinforcement learning eviction policy integrated into NGINX with strict SLOs.

Beamforming techniques are considered as essential parts to compensate the severe path loss in millimeter-wave (mmWave) communications by adopting large antenna arrays and formulating narrow beams to obtain satisfactory received powers. However, performing accurate beam alignment over such narrow beams for efficient link configuration by traditional beam selection approaches, mainly relied on channel state information, typically impose significant latency and computing overheads, which is often infeasible in vehicle-to-vehicle (V2V) communications like highly dynamic scenarios. In contrast, utilizing out-of-band contextual information, such as vehicular position information, is a potential alternative to reduce such overheads. In this context, this paper presents a deep learning-based solution on utilizing the vehicular position information for predicting the optimal beams having sufficient mmWave received powers so that the best V2V line-of-sight links can be ensured proactively. After experimental evaluation of the proposed solution on real-world measured mmWave sensing and communications datasets, the results show that the solution can achieve up to 84.58% of received power of link status on average, which confirm a promising solution for beamforming in mmWave at 60 GHz enabled V2V communications.

Despite recent advances, efficient and robust turn-taking detection remains a significant challenge in industrial-grade Voice AI agent deployments. Many existing systems rely solely on acoustic or semantic cues, leading to suboptimal accuracy and stability, while recent attempts to endow large language models with full-duplex capabilities require costly full-duplex data and incur substantial training and deployment overheads, limiting real-time performance. In this paper, we propose JAL-Turn, a lightweight and efficient speech-only turn-taking framework that adopts a joint acoustic-linguistic modeling paradigm, in which a cross-attention module adaptively integrates pre-trained acoustic representations with linguistic features to support low-latency prediction of hold vs shift states. By sharing a frozen ASR encoder, JAL-Turn enables turn-taking prediction to run fully in parallel with speech recognition, introducing no additional end-to-end latency or computational overhead. In addition, we introduce a scalable data construction pipeline that automatically derives reliable turn-taking labels from large-scale real-world dialogue corpora. Extensive experiments on public multilingual benchmarks and an in-house Japanese customer-service dataset show that JAL-Turn consistently outperforms strong state-of-the-art baselines in detection accuracy while maintaining superior real-time performance.

Reinforcement learning (RL) has enhanced the capabilities of large language models (LLMs) through reward-driven training. Nevertheless, this process can introduce excessively long responses, inflating inference latency and computational overhead. Prior length-control approaches typically rely on fixed heuristic reward shaping, which can misalign with the task objective and require brittle tuning. In this work, we propose LACONIC, a reinforcement learning method that enforces a target token budget during training. Specifically, we update policy models using an augmented objective that combines the task reward with a length-based cost. To balance brevity and task performance, the cost scale is adaptively adjusted throughout training. This yields robust length control while preserving task reward. We provide a theoretical guarantee that support the method. Across mathematical reasoning models and datasets, LACONIC preserves or improves pass@1 while reducing output length by over 50%. It maintains out-of-domain performance on general knowledge and multilingual benchmarks with 44% fewer tokens. Moreover, LACONIC integrates into standard RL-tuning with no inference changes and minimal deployment overhead.

While Large Reasoning Models (LRMs) have demonstrated impressive capabilities in solving complex tasks through the generation of long reasoning chains, this reliance on verbose generation results in significant latency and computational overhead. To address these challenges, we propose CoSMo (Consistency-Guided Split-Merge Optimization), a framework designed to eliminate structural redundancy rather than indiscriminately restricting token volume. Specifically, CoSMo utilizes a split-merge algorithm that dynamically refines reasoning chains by merging redundant segments and splitting logical gaps to ensure coherence. We then employ structure-aligned reinforcement learning with a novel segment-level budget to supervise the model in maintaining efficient reasoning structures throughout training. Extensive experiments across multiple benchmarks and backbones demonstrate that CoSMo achieves superior performance, improving accuracy by 3.3 points while reducing segment usage by 28.7\% on average compared to reasoning efficiency baselines.

Approximate Nearest Neighbor Search (ANNS), as the core of vector databases (VectorDBs), has become widely used in modern AI and ML systems, powering applications from information retrieval to bio-informatics. While graph-based ANNS methods achieve high query efficiency, their scalability is constrained by the available host memory. Recent disk-based ANNS approaches mitigate memory usage by offloading data to Solid-State Drives (SSDs). However, they still suffer from issues such as long I/O traversal path, misalignment with storage I/O granularity, and high in-memory indexing overhead, leading to significant I/O latency and ultimately limiting scalability for large-scale vector search. In this paper, we propose PageANN, a disk-based approximate nearest neighbor search (ANNS) framework designed for high performance and scalability. PageANN introduces a page-node graph structure that aligns logical graph nodes with physical SSD pages, thereby shortening I/O traversal paths and reducing I/O operations. Specifically, similar vectors are clustered into page nodes, and a co-designed disk data layout leverages this structure with a merging technique to store only representative vectors and topology information, avoiding unnecessary reads. To further improve efficiency, we design a memory management strategy that combines lightweight indexing with coordinated memory-disk data allocation, maximizing host memory utilization while minimizing query latency and storage overhead. Experimental results show that PageANN significantly outperforms state-of-the-art (SOTA) disk-based ANNS methods, achieving 1.85x-10.83x higher throughput and 51.7%-91.9% lower latency across different datasets and memory budgets, while maintaining comparable high recall accuracy.

Cybersecurity honeypots are deception tools for engaging attackers and gather intelligence, but traditional low or medium-interaction honeypots often rely on static, pre-scripted interactions that can be easily identified by skilled adversaries. This Report presents LLMHoney, an SSH honeypot that leverages Large Language Models (LLMs) to generate realistic, dynamic command outputs in real time. LLMHoney integrates a dictionary-based virtual file system to handle common commands with low latency while using LLMs for novel inputs, achieving a balance between authenticity and performance. We implemented LLMHoney using open-source LLMs and evaluated it on a testbed with 138 representative Linux commands. We report comprehensive metrics including accuracy (exact-match, Cosine Similarity, Jaro-Winkler Similarity, Levenshtein Similarity and BLEU score), response latency and memory overhead. We evaluate LLMHoney using multiple LLM backends ranging from 0.36B to 3.8B parameters, including both open-source models and a proprietary model(Gemini). Our experiments compare 13 different LLM variants; results show that Gemini-2.0 and moderately-sized models Qwen2.5:1.5B and Phi3:3.8B provide the most reliable and accurate responses, with mean latencies around 3 seconds, whereas smaller models often produce incorrect or out-of-character outputs. We also discuss how LLM integration improves honeypot realism and adaptability compared to traditional honeypots, as well as challenges such as occasional hallucinated outputs and increased resource usage. Our findings demonstrate that LLM-driven honeypots are a promising approach to enhance attacker engagement and collect richer threat intelligence.

Large language models (LLMs) benefit from test-time scaling but are often hampered by high inference latency. Speculative decoding is a natural way to accelerate the scaling process; however, scaling along both the parallel and sequential dimensions poses significant challenges, including substantial memory-bound execution and synchronization overhead. We introduce ATTS (Asynchronous Test-Time Scaling), a statistically guaranteed adaptive scaling framework that follows the hypothesis testing process to address these challenges. By revisiting arithmetic intensity, ATTS identifies synchronization as the primary bottleneck. It enables asynchronous inference through online calibration and proposes an ordinal classification algorithm that supports a three-stage rejection sampling pipeline, scaling along both the sequential and parallel axes. Across experiments on the MATH, AMC23, AIME24, and AIME25 datasets and across multiple draft-target model families, we show that ATTS delivers up to 56.7x speedup in test-time scaling and a 4.14x throughput improvement, while maintaining accurate control of the rejection rate, reducing latency and memory overhead, and incurring no accuracy loss. By scaling both in parallel and sequential dimensions, we enable the 1.5B/70B draft/target model combination to achieve the performance of the state-of-the-art reasoning model o3-mini (high) on the AIME dataset. We have released the code at https://github.com/menik1126/asynchronous-test-time-scaling.

Agentic LLM frameworks that rely on prompted orchestration, where the model itself determines workflow transitions, often suffer from hallucinated routing, infinite loops, and non-reproducible execution. We introduce GraphBit, an engine-orchestrated framework that defines workflows explicitly and deterministically as a directed acyclic graph (DAG). Unlike prompted orchestration, agents in GraphBit operate as typed functions, while a Rust-based engine governs routing, state transitions, and tool invocation, ensuring reproducibility and auditability. The engine supports parallel branch execution, conditional control flow over structured state predicates, and configurable error recovery. A three-tier memory architecture consisting of ephemeral scratch space, structured state, and external connectors isolates context across stages, preventing cascading context bloat that degrades reasoning in long-running pipelines. Across GAIA benchmark tasks spanning zero-tool, document-augmented, and web-enabled workflows, GraphBit outperforms six existing frameworks, achieving the highest accuracy (67.6 percent), zero framework-induced hallucinations, the lowest latency (11.9 ms overhead), and the highest throughput. Ablation studies demonstrate that each memory tier contributes measurably to performance, with deterministic execution providing the greatest gains on tool-intensive tasks representative of real-world deployments.

Parameter-efficient finetuning (PEFT) is a widely used technique to adapt large language models for different tasks. Service providers typically create separate systems for users to perform PEFT model finetuning and inference tasks. This is because existing systems cannot handle workloads that include a mix of inference and PEFT finetuning requests. As a result, shared GPU resources are underutilized, leading to inefficiencies. To address this problem, we present FlexLLM, the first system that can serve inference and parameter-efficient finetuning requests in the same iteration. Our system leverages the complementary nature of these two tasks and utilizes shared GPU resources to run them jointly, using a method called co-serving. To achieve this, FlexLLM introduces a novel token-level finetuning mechanism, which breaks down the finetuning computation of a sequence into smaller token-level computations and uses dependent parallelization and graph pruning, two static compilation optimizations, to minimize the memory overhead and latency for co-serving. Compared to existing systems, FlexLLM's co-serving approach reduces the activation GPU memory overhead by up to 8x, and the end-to-end GPU memory requirement of finetuning by up to 36% while maintaining a low inference latency and improving finetuning throughput. For example, under a heavy inference workload, FlexLLM can still preserve more than 80% of the peak finetuning throughput, whereas existing systems cannot make any progress with finetuning. The source code of FlexLLM is publicly available at https://github.com/flexflow/FlexFlow.

Scheduling policies in large-scale Automatic Speech Recognition (ASR) serving pipelines play a key role in determining end-to-end (E2E) latency. Yet, widely used serving engines rely on first-come-first-served (FCFS) scheduling, which ignores variability in request duration and leads to head-of-line blocking under workload drift. We show that audio duration is an accurate proxy for job processing time in ASR models such as Whisper, and use this insight to enable duration-aware scheduling. We integrate two classical algorithms, Shortest Job First (SJF) and Highest Response Ratio Next (HRRN), into vLLM and evaluate them under realistic and drifted workloads. On LibriSpeech test-clean, compared to baseline, SJF reduces median E2E latency by up to 73% at high load, but increases 90th-percentile tail latency by up to 97% due to starvation of long requests. HRRN addresses this trade-off: it reduces median E2E latency by up to 28% while bounding tail-latency degradation to at most 24%. These gains persist under workload drift, with no throughput penalty and <0.1\,ms scheduling overhead per request.

Large language models excel at reasoning but lack key aspects of introspection, including anticipating their own success and the computation required to achieve it. Humans use real-time introspection to decide how much effort to invest, when to make multiple attempts, when to stop, and when to signal success or failure. Without this, LLMs struggle to make intelligent meta-cognition decisions. Test-time scaling methods like Best-of-N drive up cost and latency by using a fixed budget of samples regardless of the marginal benefit of each one at any point in generation, and the absence of confidence signals can mislead people, prevent appropriate escalation to better tools, and undermine trustworthiness. Learned verifiers or reward models can provide confidence estimates, but do not enable adaptive inference and add substantial cost by requiring extra models or forward passes. We present ZIP-RC, which equips models with zero-overhead introspective predictions of reward and cost. At every token, ZIP-RC reuses reserved or unused logits in the same forward pass as next-token prediction to output a joint distribution over final reward and remaining length -- no extra models, architecture change, or inference overhead. This full joint distribution is used to compute a sampling utility which is the linear combination of the expected maximum reward, total compute, and latency of set of samples if generated to completion. During inference, we maximize this utility with meta-actions that determine which prefix of tokens to continue or initiate sampling from. On mixed-difficulty mathematical benchmarks, ZIP-RC improves accuracy by up to 12% over majority voting at equal or lower average cost, and traces smooth Pareto frontiers between quality, compute, and latency. By providing real-time reward-cost introspection, ZIP-RC enables adaptive, efficient reasoning.

With the wide adoption of language models for IR -- and specifically RAG systems -- the latency of the underlying LLM becomes a crucial bottleneck, since the long contexts of retrieved passages lead large prompts and therefore, compute increase. Prompt compression, which reduces the size of input prompts while aiming to preserve performance on downstream tasks, has established itself as a cost-effective and low-latency method for accelerating inference in large language models. However, its usefulness depends on whether the additional preprocessing time during generation is offset by faster decoding. We present the first systematic, large-scale study of this trade-off, with thousands of runs and 30,000 queries across several open-source LLMs and three GPU classes. Our evaluation separates compression overhead from decoding latency while tracking output quality and memory usage. LLMLingua achieves up to 18% end-to-end speed-ups, when prompt length, compression ratio, and hardware capacity are well matched, with response quality remaining statistically unchanged across summarization, code generation, and question answering tasks. Outside this operating window, however, the compression step dominates and cancels out the gains. We also show that effective compression can reduce memory usage enough to offload workloads from data center GPUs to commodity cards, with only a 0.3s increase in latency. Our open-source profiler predicts the latency break-even point for each model-hardware setup, providing practical guidance on when prompt compression delivers real-world benefits.

Machine learning (ML) inference platforms are tasked with balancing two competing goals: ensuring high throughput given many requests, and delivering low-latency responses to support interactive applications. Unfortunately, existing platform knobs (e.g., batch sizes) fail to ease this fundamental tension, and instead only enable users to harshly trade off one property for the other. This paper explores an alternate strategy to taming throughput-latency tradeoffs by changing the granularity at which inference is performed. We present Apparate, a system that automatically applies and manages early exits (EEs) in ML models, whereby certain inputs can exit with results at intermediate layers. To cope with the time-varying overhead and accuracy challenges that EEs bring, Apparate repurposes exits to provide continual feedback that powers several novel runtime monitoring and adaptation strategies. Apparate lowers median response latencies by 40.5--91.5% and 10.0--24.2% for diverse CV and NLP classification workloads, and median time-per-token latencies by 22.6--77.9% for generative scenarios, without affecting throughputs or violating tight accuracy constraints.

Large language models (LLMs) have achieved near-human performance across diverse reasoning tasks, yet their deployment on resource-constrained Internet-of-Things (IoT) devices remains impractical due to massive parameter footprints and memory-intensive autoregressive decoding. While split computing offers a promising solution by partitioning model execution between edge devices and cloud servers, existing approaches fail to address the unique challenges of autoregressive inference, particularly the iterative token generation process and expanding key-value (KV) cache requirements. This work introduces the first autoregressive-aware split computing framework designed explicitly for LLM deployment on edge devices. Our approach makes three key contributions. First, we develop one-point split compression (OPSC), a mixed-precision quantization scheme that prevents out-of-memory failures by strategically partitioning models into front-end and back-end segments with different precision levels. Second, we propose a two-stage intermediate compression pipeline that combines threshold splitting (TS) and token-wise adaptive bit quantization (TAB-Q) to preserve accuracy-critical activations while dramatically reducing communication overhead. Third, we formulate a unified optimization framework that jointly selects optimal split points, quantization settings, and sequence lengths to satisfy strict memory and latency constraints. Extensive evaluations across diverse LLMs and hardware platforms demonstrate superior performance compared to state-of-the-art quantization methods, including SmoothQuant, OmniQuant, and Atom. The framework achieves a 1.49 inference speedup and significant communication overhead reduction while maintaining or improving model accuracy.

Efficient inference for on-device Large Language Models (LLMs) remains challenging due to limited hardware resources and the high cost of the prefill stage, which processes the full input context to construct Key-Value (KV) caches. We present SparKV, an adaptive KV loading framework that combines cloud-based KV streaming with on-device computation. SparKV models the cost of individual KV chunks and decides whether each chunk should be streamed or computed locally, while overlapping the two execution paths to reduce latency. To handle fluctuations in wireless connectivity and edge resource availability, SparKV further refines offline-generated schedules at runtime to rebalance communication and computation costs. Experiments across diverse datasets, LLMs, and edge devices show that SparKV reduces Time-to-First-Token by 1.3$x-5.1x with negligible impact on response quality, while lowering per-request energy consumption by 1.5x to 3.3x, demonstrating its robustness and practicality for real-world on-device deployment.

The Mixtures-of-Experts (MoE) model is a widespread distributed and integrated learning method for large language models (LLM), which is favored due to its ability to sparsify and expand models efficiently. However, the performance of MoE is limited by load imbalance and high latency of All-To-All communication, along with relatively redundant computation owing to large expert capacity. Load imbalance may result from existing routing policies that consistently tend to select certain experts. The frequent inter-node communication in the All-To-All procedure also significantly prolongs the training time. To alleviate the above performance problems, we propose a novel routing strategy that combines load balance and locality by converting partial inter-node communication to that of intra-node. Notably, we elucidate that there is a minimum threshold for expert capacity, calculated through the maximal angular deviation between the gating weights of the experts and the assigned tokens. We port these modifications on the PanGu-Sigma model based on the MindSpore framework with multi-level routing and conduct experiments on Ascend clusters. The experiment results demonstrate that the proposed LocMoE reduces training time per epoch by 12.68% to 22.24% compared to classical routers, such as hash router and switch router, without impacting the model accuracy.

This article summarises the current status of classical communication networks and identifies some critical open research challenges that can only be solved by leveraging quantum technologies. By now, the main goal of quantum communication networks has been security. However, quantum networks can do more than just exchange secure keys or serve the needs of quantum computers. In fact, the scientific community is still investigating on the possible use cases/benefits that quantum communication networks can bring. Thus, this article aims at pointing out and clearly describing how quantum communication networks can enhance in-network distributed computing and reduce the overall end-to-end latency, beyond the intrinsic limits of classical technologies. Furthermore, we also explain how entanglement can reduce the communication complexity (overhead) that future classical virtualised networks will experience.

Real-time speech synthesis requires balancing inference latency and acoustic fidelity for interactive applications. Conventional continuous text-to-speech pipelines require computationally intensive neural vocoders to reconstruct phase information, creating a significant streaming bottleneck. Furthermore, regression-based acoustic modeling frequently induces spectral over-smoothing artifacts. To address these limitations, this paper proposes a novel end-to-end non-autoregressive architecture optimized for ultra-low latency block-wise generation, directly modeling the highly compressed discrete latent space of the Mimi neural audio codec. Integrating a modified FastSpeech 2 backbone with a progressive depth-wise sequential decoding strategy, the architecture dynamically conditions 32 layers of residual vector quantization codes. This mechanism resolves phonetic alignment degradation and manages the complexity of high-fidelity discrete representations without temporal autoregressive overhead. Experimental evaluations on English and Malay datasets validate its language-independent deployment capability. Compared to conventional continuous regression models, the proposed architecture demonstrates quantitative improvements in fundamental voicing accuracy and mitigates high-frequency spectral degradation. It achieves ultra-low latency inference, translating to a 10.6-fold absolute acceleration over conventional cascaded pipelines. Crucially, the system achieves an average time-to-first-byte latency of 48.99 milliseconds, falling significantly below the human perception threshold for real-time interactive streaming. These results firmly establish the proposed architecture as a highly optimized solution for deploying real-time streaming speech interfaces.

Large Language Models (LLMs) have achieved strong performance across natural language and multimodal tasks, yet their practical deployment remains constrained by inference latency and kernel launch overhead, particularly in interactive, short-sequence settings. This paper presents a hybrid runtime framework that combines Just-In-Time (JIT) compilation with CUDA Graph execution to reduce launch overhead while preserving runtime flexibility during autoregressive decoding. The framework partitions transformer inference into static components executed via CUDA Graph replay and dynamic components handled through JIT-compiled kernels, enabling asynchronous graph capture and reuse across decoding steps. We evaluate the proposed approach on LLaMA-2 7B using single-GPU, batch-size-one inference across prompt lengths from 10 to 500 tokens. Experimental results show that the hybrid runtime reduces Time-to-First-Token (TTFT) by up to 66.0% and achieves lower P99 latency compared with TensorRT-LLM in this regime. These results indicate that hybrid JIT-CUDA Graph execution can effectively reduce inference latency and variance for short-sequence LLM workloads, making it a practical optimization strategy for latency-sensitive AI applications.

Large language model retrievers improve performance on complex queries, but their practical value depends on efficiency, robustness, and reliable confidence signals in addition to accuracy. We reproduce a reasoning-intensive retrieval benchmark (BRIGHT) across 12 tasks and 14 retrievers, and extend evaluation with cold-start indexing cost, query latency distributions and throughput, corpus scaling, robustness to controlled query perturbations, and confidence use (AUROC) for predicting query success. We also quantify reasoning overhead by comparing standard queries to five provided reasoning-augmented variants, measuring accuracy gains relative to added latency. We find that some reasoning-specialized retrievers achieve strong effectiveness while remaining competitive in throughput, whereas several large LLM-based bi-encoders incur substantial latency for modest gains. Reasoning augmentation incurs minimal latency for sub-1B encoders but exhibits diminishing returns for top retrievers and may reduce performance on formal math/code domains. Confidence calibration is consistently weak across model families, indicating that raw retrieval scores are unreliable for downstream routing without additional calibration. We release all code and artifacts for reproducibility.

Remote photoplethysmography (rPPG), enabling non-contact physiological monitoring through facial light reflection analysis, faces critical computational bottlenecks as deep learning introduces performance gains at the cost of prohibitive resource demands. This paper proposes ME-rPPG, a memory-efficient algorithm built on temporal-spatial state space duality, which resolves the trilemma of model scalability, cross-dataset generalization, and real-time constraints. Leveraging a transferable state space, ME-rPPG efficiently captures subtle periodic variations across facial frames while maintaining minimal computational overhead, enabling training on extended video sequences and supporting low-latency inference. Achieving cross-dataset MAEs of 5.38 (MMPD), 0.70 (VitalVideo), and 0.25 (PURE), ME-rPPG outperforms all baselines with improvements ranging from 21.3% to 60.2%. Our solution enables real-time inference with only 3.6 MB memory usage and 9.46 ms latency -- surpassing existing methods by 19.5%-49.7% accuracy and 43.2% user satisfaction gains in real-world deployments. The code and demos are released for reproducibility on https://health-hci-group.github.io/ME-rPPG-demo/.

Spiking Neural Networks (SNNs) are energy efficient alternatives to commonly used deep neural networks (DNNs). Through event-driven information processing, SNNs can reduce the expensive compute requirements of DNNs considerably, while achieving comparable performance. However, high inference latency is a significant hindrance to the edge deployment of deep SNNs. Computation over multiple timesteps not only increases latency as well as overall energy budget due to higher number of operations, but also incurs memory access overhead of fetching membrane potentials, both of which lessen the energy benefits of SNNs. To overcome this bottleneck and leverage the full potential of SNNs, we propose an Iterative Initialization and Retraining method for SNNs (IIR-SNN) to perform single shot inference in the temporal axis. The method starts with an SNN trained with T timesteps (T>1). Then at each stage of latency reduction, the network trained at previous stage with higher timestep is utilized as initialization for subsequent training with lower timestep. This acts as a compression method, as the network is gradually shrunk in the temporal domain. In this paper, we use direct input encoding and choose T=5, since as per literature, it is the minimum required latency to achieve satisfactory performance on ImageNet. The proposed scheme allows us to obtain SNNs with up to unit latency, requiring a single forward pass during inference. We achieve top-1 accuracy of 93.05%, 70.15% and 67.71% on CIFAR-10, CIFAR-100 and ImageNet, respectively using VGG16, with just 1 timestep. In addition, IIR-SNNs perform inference with 5-2500X reduced latency compared to other state-of-the-art SNNs, maintaining comparable or even better accuracy. Furthermore, in comparison with standard DNNs, the proposed IIR-SNNs provide25-33X higher energy efficiency, while being comparable to them in classification performance.

Deep neural network (DNN) inference has become an important part of many data-center workloads. This has prompted focused efforts to design ever-faster deep learning accelerators such as GPUs and TPUs. However, an end-to-end DNN-based vision application contains more than just DNN inference, including input decompression, resizing, sampling, normalization, and data transfer. In this paper, we perform a thorough evaluation of computer vision inference requests performed on a throughput-optimized serving system. We quantify the performance impact of server overheads such as data movement, preprocessing, and message brokers between two DNNs producing outputs at different rates. Our empirical analysis encompasses many computer vision tasks including image classification, segmentation, detection, depth-estimation, and more complex processing pipelines with multiple DNNs. Our results consistently demonstrate that end-to-end application performance can easily be dominated by data processing and data movement functions (up to 56% of end-to-end latency in a medium-sized image, and sim 80% impact on system throughput in a large image), even though these functions have been conventionally overlooked in deep learning system design. Our work identifies important performance bottlenecks in different application scenarios, achieves 2.25times better throughput compared to prior work, and paves the way for more holistic deep learning system design.

Recent work has begun to equip vision-language-action (VLA) policies with explicit intermediate reasoning. In embodied control, however, textual chain-of-thought is a poor fit: irrelevant or weakly textual information can interfere with action prediction, while autoregressive text decoding adds too much latency for real-time closed-loop execution. We present VISUALTHINK-VLA, a visual intermediate-reasoning framework for accurate, low-latency VLA policies. Our bootstrapping philosophy is to guide action with effective visual thinking: VISUALTHINK-VLA bootstraps action prediction through a compact visual-evidence interface that preserves spatial precision while avoiding decoding overhead. Besides, to further improve performance and efficiency, VISUALTHINK-VLA adopts a tailored selective routing mechanism to learn the visual evidence tokens, enabling low-latency inference while preserving high-capacity specialization. We also introduce VisualEvidence-Kit, a supervision-and-audit resource centered on a VisualEvidence-Agent that constructs a 754.7k VLA instructions VisualEvidence-Set for route supervision and counterfactual faithfulness tests. Across multiple benchmarks and real-robot evaluation, VISUALTHINK-VLA achieves the highest success rate on most benchmarks while reducing the multi-second latency of reasoning-augmented baselines to the sub-second regime. For example, on BridgeData V2, it reduces step latency from 8.377,s with ECoT to 0.367,s, achieving a 22.8 times speedup.

Large Language Models (LLMs), such as OpenAI-o1 and DeepSeek-R1, have demonstrated strong reasoning capabilities. To further enhance LLM capabilities, recent agentic systems, such as Deep Research, incorporate web interactions into LLM reasoning to mitigate uncertainties and reduce potential errors. However, existing research predominantly focuses on reasoning performance, often neglecting the efficiency of agentic systems. In this work, we present a comprehensive empirical study that identifies efficiency bottlenecks in web-interactive agentic systems. We decompose end-to-end latency into two primary components: LLM API latency and web environment latency. We conduct a comprehensive empirical study across 15 models and 5 providers to demonstrate high variability in API-based agentic systems. We observe that web environment latency can contribute as much as 53.7% to the overall latency in a web-based agentic system. To improve latency, we propose SpecCache, a caching framework augmented with speculative execution that can reduce web environment overhead. Extensive evaluations on two standard benchmarks show that our approach improves the cache hit rate by up to 58x compared to a random caching strategy, while reducing web environment overhead by up to 3.2x, without degrading agentic system performance.

For deployment, neural architecture search should be hardware-aware, in order to satisfy the device-specific constraints (e.g., memory usage, latency and energy consumption) and enhance the model efficiency. Existing methods on hardware-aware NAS collect a large number of samples (e.g., accuracy and latency) from a target device, either builds a lookup table or a latency estimator. However, such approach is impractical in real-world scenarios as there exist numerous devices with different hardware specifications, and collecting samples from such a large number of devices will require prohibitive computational and monetary cost. To overcome such limitations, we propose Hardware-adaptive Efficient Latency Predictor (HELP), which formulates the device-specific latency estimation problem as a meta-learning problem, such that we can estimate the latency of a model's performance for a given task on an unseen device with a few samples. To this end, we introduce novel hardware embeddings to embed any devices considering them as black-box functions that output latencies, and meta-learn the hardware-adaptive latency predictor in a device-dependent manner, using the hardware embeddings. We validate the proposed HELP for its latency estimation performance on unseen platforms, on which it achieves high estimation performance with as few as 10 measurement samples, outperforming all relevant baselines. We also validate end-to-end NAS frameworks using HELP against ones without it, and show that it largely reduces the total time cost of the base NAS method, in latency-constrained settings. Code is available at https://github.com/HayeonLee/HELP.

The emergence of LLMs has catalyzed a paradigm shift in autonomous agent development, enabling systems capable of reasoning, planning, and executing complex multi-step tasks. However, existing agent frameworks often suffer from architectural rigidity, vendor lock-in, and prohibitive complexity that impedes rapid prototyping and deployment. This paper presents AgentForge, a lightweight, open-source Python framework designed to democratize the construction of LLM-driven autonomous agents through a principled modular architecture. AgentForge introduces three key innovations: (1) a composable skill abstraction that enables fine-grained task decomposition with formally defined input-output contracts, (2) a unified LLM backend interface supporting seamless switching between cloud-based APIs and local inference engines, and (3) a declarative YAML-based configuration system that separates agent logic from implementation details. We formalize the skill composition mechanism as a directed acyclic graph (DAG) and prove its expressiveness for representing arbitrary sequential and parallel task workflows. Comprehensive experimental evaluation across four benchmark scenarios demonstrates that AgentForge achieves competitive task completion rates while reducing development time by 62% compared to LangChain and 78% compared to direct API integration. Latency measurements confirm sub-100ms orchestration overhead, rendering the framework suitable for real-time applications. The modular design facilitates extension: we demonstrate the integration of six built-in skills and provide comprehensive documentation for custom skill development. AgentForge addresses a critical gap in the LLM agent ecosystem by providing researchers and practitioners with a production-ready foundation for constructing, evaluating, and deploying autonomous agents without sacrificing flexibility or performance.

Embedding-based search is widely used in applications such as recommendation and retrieval-augmented generation (RAG). Recently, there is a growing demand to support these capabilities over personal data stored locally on devices. However, maintaining the necessary data structure associated with the embedding-based search is often infeasible due to its high storage overhead. For example, indexing 100 GB of raw data requires 150 to 700 GB of storage, making local deployment impractical. Reducing this overhead while maintaining search quality and latency becomes a critical challenge. In this paper, we present LEANN, a storage-efficient approximate nearest neighbor (ANN) search index optimized for resource-constrained personal devices. LEANN combines a compact graph-based structure with an efficient on-the-fly recomputation strategy to enable fast and accurate retrieval with minimal storage overhead. Our evaluation shows that LEANN reduces index size to under 5% of the original raw data, achieving up to 50 times smaller storage than standard indexes, while maintaining 90% top-3 recall in under 2 seconds on real-world question answering benchmarks.

Deep learning models are increasingly deployed on resource-constrained edge devices for real-time data analytics. In recent years, Vision Transformer models and their variants have demonstrated outstanding performance across various computer vision tasks. However, their high computational demands and inference latency pose significant challenges for model deployment on resource-constraint edge devices. To address this issue, we propose a novel Vision Transformer splitting framework, ED-ViT, designed to execute complex models across multiple edge devices efficiently. Specifically, we partition Vision Transformer models into several sub-models, where each sub-model is tailored to handle a specific subset of data classes. To further minimize computation overhead and inference latency, we introduce a class-wise pruning technique that reduces the size of each sub-model. We conduct extensive experiments on five datasets with three model structures, demonstrating that our approach significantly reduces inference latency on edge devices and achieves a model size reduction of up to 28.9 times and 34.1 times, respectively, while maintaining test accuracy comparable to the original Vision Transformer. Additionally, we compare ED-ViT with two state-of-the-art methods that deploy CNN and SNN models on edge devices, evaluating accuracy, inference time, and overall model size. Our comprehensive evaluation underscores the effectiveness of the proposed ED-ViT framework.

Agentic multimodal large language models (MLLMs) (e.g., OpenAI o3 and Gemini Agentic Vision) achieve remarkable reasoning capabilities through iterative visual tool invocation. However, the cascaded perception, reasoning, and tool-calling loops introduce significant sequential overhead. This overhead, termed agentic depth, incurs prohibitive latency and seriously limits system-level concurrency. To this end, we propose SpecEyes, an agentic-level speculative acceleration framework that breaks this sequential bottleneck. Our key insight is that a lightweight, tool-free MLLM can serve as a speculative planner to predict the execution trajectory, enabling early termination of expensive tool chains without sacrificing accuracy. To regulate this speculative planning, we introduce a cognitive gating mechanism based on answer separability, which quantifies the model's confidence for self-verification without requiring oracle labels. Furthermore, we design a heterogeneous parallel funnel that exploits the stateless concurrency of the small model to mask the stateful serial execution of the large model, maximizing system throughput. Extensive experiments on V* Bench, HR-Bench, and POPE demonstrate that SpecEyes achieves 1.1-3.35x speedup over the agentic baseline while preserving or even improving accuracy (up to +6.7%), thereby boosting serving throughput under concurrent workloads.

Retrieval-Augmented Generation (RAG) has emerged as the predominant paradigm for grounding Large Language Model outputs in factual knowledge, effectively mitigating hallucinations. However, conventional RAG systems operate under a "retrieve-always" assumption, querying vector databases for every input regardless of query complexity. This static approach incurs substantial computational overhead and inference latency, particularly problematic for high-throughput production deployments. We introduce L-RAG (Lazy Retrieval-Augmented Generation), an adaptive framework that implements hierarchical context management through entropy-based gating. L-RAG employs a two-tier architecture: queries are first processed with a compact document summary, and expensive chunk retrieval is triggered only when the model's predictive entropy exceeds a calibrated threshold, signaling genuine uncertainty. Through experiments on SQuAD 2.0 (N=500) using the Phi-2 model, we demonstrate that L-RAG provides a tunable accuracy-efficiency trade-off: at a conservative threshold (tau=0.5), L-RAG achieves 78.2% accuracy, matching Standard RAG (77.8%), with 8% retrieval reduction; at a balanced threshold (tau=1.0), retrieval reduction increases to 26% with modest accuracy trade-off (76.0%). Latency analysis shows that L-RAG saves 80-210ms per query when retrieval latency exceeds 500ms. Analysis of entropy distributions reveals statistically significant separation (p < 0.001) between correct predictions (H=1.72) and errors (H=2.20), validating entropy as a reliable uncertainty signal. L-RAG offers a practical, training-free approach toward more efficient RAG deployment, providing system architects with a configurable knob to balance accuracy and throughput requirements.

In the context of Machine Learning as a Service (MLaaS) clouds, the extensive use of Large Language Models (LLMs) often requires efficient management of significant query loads. When providing real-time inference services, several challenges arise. Firstly, increasing the number of GPUs may lead to a decrease in inference speed due to heightened communication overhead, while an inadequate number of GPUs can lead to out-of-memory errors. Secondly, different deployment strategies need to be evaluated to guarantee optimal utilization and minimal inference latency. Lastly, inefficient orchestration of inference queries can easily lead to significant Service Level Objective (SLO) violations. Lastly, inefficient orchestration of inference queries can easily lead to significant Service Level Objective (SLO) violations. To address these challenges, we propose a Unified and Efficient approach for Large Language Model inference serving (UELLM), which consists of three main components: 1) resource profiler, 2) batch scheduler, and 3) LLM deployer. UELLM minimizes resource overhead, reduces inference latency, and lowers SLO violation rates. Compared with state-of-the-art (SOTA) techniques, UELLM reduces the inference latency by 72.3% to 90.3%, enhances GPU utilization by 1.2X to 4.1X, and increases throughput by 1.92X to 4.98X, it can also serve without violating the inference latency SLO.

Hybrid storage systems (HSS) integrate multiple storage devices with diverse characteristics to deliver high performance and capacity at low cost. The performance of an HSS highly depends on the effectiveness of two key policies: (1) the data-placement policy, which determines the best-fit storage device for incoming data, and (2) the data-migration policy, which dynamically rearranges stored data (i.e., prefetches hot data and evicts cold data) across the devices to sustain high HSS performance. Prior works optimize either data placement or data migration in isolation, which leads to suboptimal HSS performance. Unfortunately, no prior work tries to optimize both policies together. Our goal is to design a holistic data-management technique that optimizes both data-placement and data-migration policies to fully exploit the potential of an HSS, and thus significantly improve system performance. We propose Harmonia, a multi-agent reinforcement learning (RL)-based data-management technique that employs two lightweight autonomous RL agents, a data-placement agent and a data-migration agent, that adapt their policies for the current workload and HSS configuration while coordinating with each other to improve overall HSS performance. We evaluate Harmonia on real HSS configurations with up to four heterogeneous storage devices and seventeen data-intensive workloads. On performance-optimized (cost-optimized) HSS with two storage devices, Harmonia outperforms the best-performing prior approach by 49.5% (31.7%) on average. On an HSS with three (four) devices, Harmonia outperforms the best-performing prior work by 37.0% (42.0%) on average. Harmonia's performance benefits come with low latency (240ns for inference) and storage overheads (206 KiB in DRAM for both RL agents combined). We will open-source Harmonia's implementation to aid future research on HSS.

Large Language Models (LLMs) for Generative AI have achieved remarkable progress, evolving into sophisticated and versatile tools widely adopted across various domains and applications. However, the substantial memory overhead caused by their vast number of parameters, combined with the high computational demands of the attention mechanism, poses significant challenges in achieving low latency and high throughput for LLM inference services. Recent advancements, driven by groundbreaking research, have significantly accelerated progress in this field. This paper provides a comprehensive survey of these methods, covering fundamental instance-level approaches, in-depth cluster-level strategies, emerging scenario directions, and other miscellaneous but important areas. At the instance level, we review model placement, request scheduling, decoding length prediction, storage management, and the disaggregation paradigm. At the cluster level, we explore GPU cluster deployment, multi-instance load balancing, and cloud service solutions. For emerging scenarios, we organize the discussion around specific tasks, modules, and auxiliary methods. To ensure a holistic overview, we also highlight several niche yet critical areas. Finally, we outline potential research directions to further advance the field of LLM inference serving.

The platonic representation hypothesis suggests that sufficiently large models converge to a shared representation geometry, even across modalities. Motivated by this, we ask: Can the semantic knowledge of a language model efficiently improve a vision model? As an answer, we introduce TextTeacher, a simple auxiliary objective that injects text embeddings as additional information into image classification training. TextTeacher uses readily available image captions, a pre-trained and frozen text encoder, and a lightweight projection to produce semantic anchors that efficiently guide representations during training while leaving the inference-time model unchanged. On ImageNet with standard ViT backbones, TextTeacher improves accuracy by up to +2.7 percentage points (p.p.) and yields consistent transfer gains (on average +1.0 p.p.) under the same recipe and compute. It outperforms vision knowledge distillation, yielding more accuracy at a constant compute budget or similar accuracy, but 33% faster. Our analysis indicates that TextTeacher acts as a feature-space preconditioner, shaping deeper layers in the first stages of training, and aiding generalization by supplying complementary semantic cues. TextTeacher adds negligible overhead, requires no costly multimodal training of the target model and preserves the simplicity and latency of pure vision models. Project page with code and captions: https://nauen-it.de/publications/text-teacher

Private BitTorrent trackers enforce upload-to-download ratios to prevent free-riding, but suffer from three critical weaknesses: reputation cannot move between trackers, centralized servers create single points of failure, and upload statistics are self-reported and unverifiable. When a tracker shuts down, users lose their contribution history and cannot prove their standing to new communities. We address these problems by storing reputation in smart contracts and replacing self-reports with cryptographic attestations. Peers sign receipts for received pieces; the tracker aggregates them via BLS signatures and updates reputation. If a tracker is unavailable, peers fall back to an authenticated distributed hash table (DHT): stored reputation acts as a public key infrastructure (PKI), preserving access control without the tracker. Reputation is portable across tracker failures through single-hop migration in factory-deployed contracts. We also address the privacy implications of publishing public keys and reputations tied to private trackers on a public ledger: we propose ephemeral session keys to prevent linking peer identities, zero-knowledge membership proofs for anonymous DHT participation, and confidential reputation using homomorphic commitments. We formalize the security requirements, prove four security properties under standard cryptographic assumptions, and evaluate a prototype. Measurements show that transfer receipts add less than 5\% end-to-end overhead with typical piece sizes. To minimize signing overhead, we adopt a hybrid signature scheme: ECDSA signs individual piece receipts at transfer time for low per-operation latency, while BLS serves as the overarching scheme, enabling compact aggregation of many receipts into a single proof at report time. This design reduces client-side signing cost by an order of magnitude compared to using BLS throughout.

Confidential computing on GPUs, like NVIDIA H100, mitigates the security risks of outsourced Large Language Models (LLMs) by implementing strong isolation and data encryption. Nonetheless, this encryption incurs a significant performance overhead, reaching up to 52.8 percent and 88.2 percent throughput drop when serving OPT-30B and OPT-66B, respectively. To address this challenge, we introduce PipeLLM, a user-transparent runtime system. PipeLLM removes the overhead by overlapping the encryption and GPU computation through pipelining - an idea inspired by the CPU instruction pipelining - thereby effectively concealing the latency increase caused by encryption. The primary technical challenge is that, unlike CPUs, the encryption module lacks prior knowledge of the specific data needing encryption until it is requested by the GPUs. To this end, we propose speculative pipelined encryption to predict the data requiring encryption by analyzing the serving patterns of LLMs. Further, we have developed an efficient, low-cost pipeline relinquishing approach for instances of incorrect predictions. Our experiments on NVIDIA H100 GPU show that compared with vanilla systems without confidential computing (e.g., vLLM, PEFT, and FlexGen), PipeLLM incurs modest overhead (less than 19.6 percent in throughput) across various LLM sizes, from 13B to 175B.

Chain-of-Thought (CoT) prompting improves reasoning but often produces long and redundant traces that substantially increase inference cost. We present SyncThink, a training-free and plug-and-play decoding method that reduces CoT overhead without modifying model weights. We find that answer tokens attend weakly to early reasoning and instead focus on the special token "/think", indicating an information bottleneck. Building on this observation, SyncThink monitors the model's own reasoning-transition signal and terminates reasoning. Experiments on GSM8K, MMLU, GPQA, and BBH across three DeepSeek-R1 distilled models show that SyncThink achieves 62.00 percent average Top-1 accuracy using 656 generated tokens and 28.68 s latency, compared to 61.22 percent, 2141 tokens, and 92.01 s for full CoT decoding. On long-horizon tasks such as GPQA, SyncThink can further yield up to +8.1 absolute accuracy by preventing over-thinking.

Graph Retrieval-Augmented Generation (GraphRAG) has emerged as a key technique for enhancing Large Language Models (LLMs) with proprietary Knowledge Graphs (KGs) in knowledge-intensive applications. As these KGs often represent an organization's highly valuable intellectual property (IP), they face a significant risk of theft for private use. In this scenario, attackers operate in isolated environments. This private-use threat renders passive defenses like watermarking ineffective, as they require output access for detection. Simultaneously, the low-latency demands of GraphRAG make strong encryption which incurs prohibitive overhead impractical. To address these challenges, we propose AURA, a novel framework based on Data Adulteration designed to make any stolen KG unusable to an adversary. Our framework pre-emptively injects plausible but false adulterants into the KG. For an attacker, these adulterants deteriorate the retrieved context and lead to factually incorrect responses. Conversely, for authorized users, a secret key enables the efficient filtering of all adulterants via encrypted metadata tags before they are passed to the LLM, ensuring query results remain completely accurate. Our evaluation demonstrates the effectiveness of this approach: AURA degrades the performance of unauthorized systems to an accuracy of just 5.3%, while maintaining 100% fidelity for authorized users with negligible overhead. Furthermore, AURA proves robust against various sanitization attempts, retaining 80.2% of its adulterants.

REPA and its variants effectively mitigate training challenges in diffusion models by incorporating external visual representations from pretrained models, through alignment between the noisy hidden projections of denoising networks and foundational clean image representations. We argue that the external alignment, which is absent during the entire denoising inference process, falls short of fully harnessing the potential of discriminative representations. In this work, we propose a straightforward method called Representation Entanglement for Generation (REG), which entangles low-level image latents with a single high-level class token from pretrained foundation models for denoising. REG acquires the capability to produce coherent image-class pairs directly from pure noise, substantially improving both generation quality and training efficiency. This is accomplished with negligible additional inference overhead, requiring only one single additional token for denoising (<0.5\% increase in FLOPs and latency). The inference process concurrently reconstructs both image latents and their corresponding global semantics, where the acquired semantic knowledge actively guides and enhances the image generation process. On ImageNet 256times256, SiT-XL/2 + REG demonstrates remarkable convergence acceleration, achieving 63times and 23times faster training than SiT-XL/2 and SiT-XL/2 + REPA, respectively. More impressively, SiT-L/2 + REG trained for merely 400K iterations outperforms SiT-XL/2 + REPA trained for 4M iterations (10times longer). Code is available at: https://github.com/Martinser/REG.

The increasing size of large language models (LLMs) challenges their usage on resource-constrained platforms. For example, memory on modern GPUs is insufficient to hold LLMs that are hundreds of Gigabytes in size. Offloading is a popular method to escape this constraint by storing weights of an LLM model to host CPU memory and SSD, then loading each weight to GPU before every use. In our case study of offloaded inference, we found that due to the low bandwidth between storage devices and GPU, the latency of transferring large model weights from its offloaded location to GPU memory becomes the critical bottleneck with actual compute taking nearly 0% of runtime. To effectively reduce the weight transfer latency, we propose a novel sparse format that compresses the unstructured sparse pattern of pruned LLM weights to non-zero values with high compression ratio and low decompression overhead. Endor achieves this by expressing the positions of non-zero elements with a bitmap. Compared to offloaded inference using the popular Huggingface Accelerate, applying Endor accelerates OPT-66B by 1.70x and Llama2-70B by 1.78x. When direct weight transfer from SSD to GPU is leveraged, Endor achieves 2.25x speedup on OPT-66B and 2.37x speedup on Llama2-70B.

Large Language Models (LLMs) are increasingly deployed in both latency-sensitive online services and cost-sensitive offline workloads. Co-locating these workloads on shared serving instances can improve resource utilization, but directly applying this approach to Prefill/Decode (P/D) disaggregated systems introduces severe load imbalance, as fluctuating request mixes alter the intrinsic P/D ratio. Existing dynamic adjustment techniques cannot keep up with the bursty traffic patterns of online services. We propose a latency-constraint disaggregated architecture, which separates cluster resources into latency-strict and latency-relaxed pools based on task latency requirements. This design enables flexible placement of offline decode tasks, mitigating P/D imbalance while preserving online performance. To fully exploit this flexibility, we propose (1) a bottleneck-based scheduler guided by a Roofline-based performance model for performance bottleneck based scheduling, and (2) a fast preemption mechanism that strictly enforces Service Level Objectives (SLOs) for online requests. Experiments on real-world traces show that compared to existing offline system approaches, our method improves offline throughput by up to 3x, while maintaining online request SLOs.

The inference process in Large Language Models (LLMs) is often limited due to the absence of parallelism in the auto-regressive decoding process, resulting in most operations being restricted by the memory bandwidth of accelerators. While methods such as speculative decoding have been suggested to address this issue, their implementation is impeded by the challenges associated with acquiring and maintaining a separate draft model. In this paper, we present Medusa, an efficient method that augments LLM inference by adding extra decoding heads to predict multiple subsequent tokens in parallel. Using a tree-based attention mechanism, Medusa constructs multiple candidate continuations and verifies them simultaneously in each decoding step. By leveraging parallel processing, Medusa introduces only minimal overhead in terms of single-step latency while substantially reducing the number of decoding steps required. We present two levels of fine-tuning procedures for Medusa to meet the needs of different use cases: Medusa-1: Medusa is directly fine-tuned on top of a frozen backbone LLM, enabling lossless inference acceleration. Medusa-2: Medusa is fine-tuned together with the backbone LLM, enabling better prediction accuracy of Medusa heads and higher speedup but needing a special training recipe that preserves the backbone model's capabilities. Moreover, we propose several extensions that improve or expand the utility of Medusa, including a self-distillation to handle situations where no training data is available and a typical acceptance scheme to boost the acceptance rate while maintaining generation quality. We evaluate Medusa on models of various sizes and training procedures. Our experiments demonstrate that Medusa-1 can achieve over 2.2x speedup without compromising generation quality, while Medusa-2 further improves the speedup to 2.3-3.6x.

Sparsely-activated Mixture-of-Experts (SMoE) models offer efficient pre-training and low latency but their large parameter counts create significant memory overhead, motivating research into expert compression. Contrary to recent findings favouring expert merging on discriminative benchmarks, we demonstrate that expert pruning is a superior strategy for generative tasks. We prove that merging introduces an irreducible error by causing a "functional subspace collapse", due to the loss of the router's independent, input-dependent control over experts. Leveraging this insight, we propose Router-weighted Expert Activation Pruning (REAP), a novel pruning criterion that considers both router gate-values and expert activation norms. Across a diverse set of SMoE models ranging from 20B to 1T parameters, REAP consistently outperforms merging and other pruning methods on generative benchmarks, especially at 50% compression. Notably, our method achieves near-lossless compression on code generation and tool-calling tasks with Qwen3-Coder-480B and Kimi-K2, even after pruning 50% of experts.

Recent Large Reasoning Models (LRMs) have shown substantially improved reasoning capabilities over traditional Large Language Models (LLMs) by incorporating extended thinking processes prior to producing final responses. However, excessively lengthy thinking introduces substantial overhead in terms of token consumption and latency, which is particularly unnecessary for simple queries. In this work, we introduce Large Hybrid-Reasoning Models (LHRMs), the first kind of model capable of adaptively determining whether to perform thinking based on the contextual information of user queries. To achieve this, we propose a two-stage training pipeline comprising Hybrid Fine-Tuning (HFT) as a cold start, followed by online reinforcement learning with the proposed Hybrid Group Policy Optimization (HGPO) to implicitly learn to select the appropriate thinking mode. Furthermore, we introduce a metric called Hybrid Accuracy to quantitatively assess the model's capability for hybrid thinking. Extensive experimental results show that LHRMs can adaptively perform hybrid thinking on queries of varying difficulty and type. It outperforms existing LRMs and LLMs in reasoning and general capabilities while significantly improving efficiency. Together, our work advocates for a reconsideration of the appropriate use of extended thinking processes and provides a solid starting point for building hybrid thinking systems.

Diffusion Transformers (DiT) have become the de-facto model for generating high-quality visual content like videos and images. A huge bottleneck is the attention mechanism where complexity scales quadratically with resolution and video length. One logical way to lessen this burden is sparse attention, where only a subset of tokens or patches are included in the calculation. However, existing techniques fail to preserve visual quality at extremely high sparsity levels and might even incur non-negligible compute overheads. % To address this concern, we propose Re-ttention, which implements very high sparse attention for visual generation models by leveraging the temporal redundancy of Diffusion Models to overcome the probabilistic normalization shift within the attention mechanism. Specifically, Re-ttention reshapes attention scores based on the prior softmax distribution history in order to preserve the visual quality of the full quadratic attention at very high sparsity levels. % Experimental results on T2V/T2I models such as CogVideoX and the PixArt DiTs demonstrate that Re-ttention requires as few as 3.1\% of the tokens during inference, outperforming contemporary methods like FastDiTAttn, Sparse VideoGen and MInference. Further, we measure latency to show that our method can attain over 45\% end-to-end % and over 92\% self-attention latency reduction on an H100 GPU at negligible overhead cost. Code available online here: https://github.com/cccrrrccc/Re-ttention{https://github.com/cccrrrccc/Re-ttention}

Image matching is a key component of modern 3D vision algorithms, essential for accurate scene reconstruction and localization. MASt3R redefines image matching as a 3D task by leveraging DUSt3R and introducing a fast reciprocal matching scheme that accelerates matching by orders of magnitude while preserving theoretical guarantees. This approach has gained strong traction, with DUSt3R and MASt3R collectively cited over 250 times in a short span, underscoring their impact. However, despite its accuracy, MASt3R's inference speed remains a bottleneck. On an A40 GPU, latency per image pair is 198.16 ms, mainly due to computational overhead from the ViT encoder-decoder and Fast Reciprocal Nearest Neighbor (FastNN) matching. To address this, we introduce Speedy MASt3R, a post-training optimization framework that enhances inference efficiency while maintaining accuracy. It integrates multiple optimization techniques, including FlashMatch-an approach leveraging FlashAttention v2 with tiling strategies for improved efficiency, computation graph optimization via layer and tensor fusion having kernel auto-tuning with TensorRT (GraphFusion), and a streamlined FastNN pipeline that reduces memory access time from quadratic to linear while accelerating block-wise correlation scoring through vectorized computation (FastNN-Lite). Additionally, it employs mixed-precision inference with FP16/FP32 hybrid computations (HybridCast), achieving speedup while preserving numerical precision. Evaluated on Aachen Day-Night, InLoc, 7-Scenes, ScanNet1500, and MegaDepth1500, Speedy MASt3R achieves a 54% reduction in inference time (198 ms to 91 ms per image pair) without sacrificing accuracy. This advancement enables real-time 3D understanding, benefiting applications like mixed reality navigation and large-scale 3D scene reconstruction.

Large language models (LLMs) with extended context windows have become increasingly prevalent for tackling complex tasks. However, the substantial Key-Value (KV) cache required for long-context LLMs poses significant deployment challenges. Existing approaches either discard potentially critical information needed for future generations or offer limited efficiency gains due to high computational overhead. In this paper, we introduce Chelsea, a simple yet effective framework for online KV cache clustering. Our approach is based on the observation that key states exhibit high similarity along the sequence dimension. To enable efficient clustering, we divide the sequence into chunks and propose Chunked Soft Matching, which employs an alternating partition strategy within each chunk and identifies clusters based on similarity. Chelsea then merges the KV cache within each cluster into a single centroid. Additionally, we provide a theoretical analysis of the computational complexity and the optimality of the intra-chunk partitioning strategy. Extensive experiments across various models and long-context benchmarks demonstrate that Chelsea achieves up to 80% reduction in KV cache memory usage while maintaining comparable model performance. Moreover, with minimal computational overhead, Chelsea accelerates the decoding stage of inference by up to 3.19times and reduces end-to-end latency by up to 2.72times.

Existing large language model (LLM) serving systems fall into two categories: 1) a unified system where prefill phase and decode phase are co-located on the same GPU, sharing the unified computational resource and storage, and 2) a disaggregated system where the two phases are disaggregated to different GPUs. The design of the disaggregated system addresses the latency interference and sophisticated scheduling issues in the unified system but leads to storage challenges including 1) replicated weights for both phases that prevent flexible deployment, 2) KV cache transfer overhead between the two phases, 3) storage imbalance that causes substantial wasted space of the GPU capacity, and 4) suboptimal resource adjustment arising from the difficulties in migrating KV cache. Such storage inefficiency delivers poor serving performance under high request rates. In this paper, we identify that the advantage of the disaggregated system lies in the disaggregated computation, i.e., partitioning the computational resource to enable the asynchronous computation of two phases. Thus, we propose a novel LLM serving system, semi-PD, characterized by disaggregated computation and unified storage. In semi-PD, we introduce a computation resource controller to achieve disaggregated computation at the streaming multi-processor (SM) level, and a unified memory manager to manage the asynchronous memory access from both phases. semi-PD has a low-overhead resource adjustment mechanism between the two phases, and a service-level objective (SLO) aware dynamic partitioning algorithm to optimize the SLO attainment. Compared to state-of-the-art systems, semi-PD maintains lower latency at higher request rates, reducing the average end-to-end latency per request by 1.27-2.58x on DeepSeek series models, and serves 1.55-1.72x more requests adhering to latency constraints on Llama series models.