推理成本

AI 服务暂时不可用，以下为来源摘要，待恢复后补全翻译：This article walks through what each technique actually does, why skipping them costs real money and real latency, and then gets hands-on with five specific methods people are running in production right now.

AI 服务暂时不可用，以下为来源摘要，待恢复后补全翻译：arXiv:2608.26496v1 Announce Type: new Abstract: Navigation in unknown environments to find unforeseen objects has become increasingly feasible with capable vision and language foundation models. However, these models also introduce non-negligible inference latency, which becomes an important concern when agents must operate continuously in the real world. Most state-of-the-art methods are still developed in synchronous simulators, where the environment waits for the agent to act and inference time is effectively free. As a result, agents are often designed around the sequential execution of perception, reasoning, and action, with little regard for time constraints. Under real-time execution, where wall-clock time counts towards the task budget, the inefficiencies of these architectures become clear. We show that recent zero-shot object navigation methods suffer consistent performance degradation under such realistic timing conditions. Motivated by this observation, we propose RTNav, a simple but effective architecture that treats inference latency, asynchronous environment stepping, and bounded compute as explicit design considerations. Evaluated on real-time variants of HM3D-v1, HM3D-v2, and HM3D-OVON, RTNav improves the success rate by up to 11% and the Success weighted by Completion Time by up to 5.1 points over prior work.

AI 服务暂时不可用，以下为来源摘要，待恢复后补全翻译：arXiv:2608.26383v1 Announce Type: new Abstract: Autonomous robots performing laboratory tasks depend on 3D reconstruction pipelines that can turn raw camera streams into actionable object representations within the latency budget of a physical control loop. Neural 3D reconstruction methods have demonstrated high-quality view synthesis, but their real-time viability across the compute platforms on which laboratory robots actually run remains poorly characterized. In this work, we present a systematic compute-platform benchmark of neural 3D reconstruction methods, evaluating NeRF and 3D Gaussian Splatting training and rendering on GPU-enabled computing devices ranging from single-board computers to server-class nodes, and place Meta's SAM3D single-image reconstruction on the same axes to quantify its latency and fidelity gap relative to per-scene optimization. Our results show that Gaussian Splatting yields higher rendering quality than NeRF at greater GPU cost, and that onboard compute is insufficient for full per-scene optimization at interactive rates. Our preliminary assessment on SAM3D indicates that it delivers plausible object geometry within seconds, but with detail mismatches that can compromise downstream manipulation. Together, these findings motivate tiered pipelines in which lightweight feed-forward reconstruction sustains the real-time perception-and-tracking loop for laboratory robots, while heavier neural reconstruction is scheduled selectively on suitable compute.

AI 服务暂时不可用，以下为来源摘要，待恢复后补全翻译：arXiv:2608.26112v1 Announce Type: new Abstract: Speculative decoding accelerates large language model inference through a draft-then-verify paradigm. Building on this, tree-structured methods improve inference by organizing proposals into multiple candidate paths, increasing the accepted length. However, existing tree-structured methods use a single drafter for all drafting steps, creating a dilemma: a smaller drafter is fast but yields lower-quality trees, whereas a larger drafter improves tree quality but suffers from high latency. To address this, we propose TreeGraft, a multi-drafter framework in which drafters of different costs jointly construct a shared draft tree. TreeGraft uses the stronger drafter to rescore candidates by updating scores assigned by the weaker drafter, reselect grafting positions, and recover promising paths left unexplored. It also integrates stronger drafter expansions non-destructively, preserving existing branches that may still be accepted by the target model. Together, these designs improve the quality of the shared draft tree. To control the drafting cost, TreeGraft introduces a lightweight scheduler distilled from an offline value system to decide when to call the stronger drafter. Across 10 model pairs and 6 benchmarks, TreeGraft outperforms the better of the two fixed single-drafter endpoint strategies by 15.1% on average, reaching a maximum gain of 26.6%. Our code is available at https://anonymous.4open.science/r/TreeGraft-E983.

AI 服务暂时不可用，以下为来源摘要，待恢复后补全翻译：Serving automatic speech recognition (ASR) models at scale is costly when each request uses only a fraction of a GPU. Learn how NVIDIA CUDA Multi-Process Service (MPS) with NVIDIA Triton Inference Server on Amazon EC2 GPU instances cuts GPU infrastructure by 75% while holding sub-second latency at 92.1 requests per second per GPU.

AI 服务暂时不可用，以下为来源摘要，待恢复后补全翻译：arXiv:2608.25140v1 Announce Type: new Abstract: Existing approaches to building line-level Arabic handwritten-text-recognition (HTR) training data either rely on fully manual annotation, which does not scale, or on automatic OCR-to-reference alignment methods not yet extended to multi-script, two-zone (main-plus-margin) manuscript layouts with a provable correctness guarantee. We present RefLAM (Reference-grounded Line Annotation for Manuscripts), a pipeline converting manuscript page images and clean transcriptions into validated, line-level ground truth without sacrificing human oversight. RefLAM couples a deep-learning page-segmentation model with a multimodal large language model (MLLM) for structured OCR and a diacritic-agnostic fuzzy alignment engine that grounds each OCR line in a contiguous span of the reference text, with a character-level confidence score in $[0,100]$. A perfect score is provably equivalent to character-for-character identity of the normalised strings (the Confidence-100 rule), verified with no counterexample across the released corpus. A reviewer can thus trust a perfect score, confirming most lines at a glance rather than retyping them, so annotation becomes triaged, with attention concentrated on uncertain alignments. Across 7 fully page-validated books we measured a 75$\times$ throughput gain over manual annotation (3,000 vs. 40 lines/hr); applying the same guarantee to 7 further books, we retained 16,533 confidence-100 main-text lines within one week, excluding sub-100 lines rather than manually correcting them. Using RefLAM, we release AraMS-28k: 14 historical Arabic manuscript books, 3,043 pages, and 27,971 main-text and 629 margin-line annotations with bounding boxes, layout labels, and insertion anchors for 191 margin entries (30.4%). We also finetune Muharaf-pretrained baselines (including HATFormer) on AraMS-28k and report CER results confirming its practical utility for downstream HTR training.

AI 服务暂时不可用，以下为来源摘要，待恢复后补全翻译：arXiv:2608.25068v1 Announce Type: new Abstract: Depth pruning removes entire Transformer blocks to reduce the inference cost of large language models, but disrupts the hidden-state distributions expected by downstream layers, leading to significant accuracy loss. We introduce SHIFT-LLM, a training-free post-pruning correction framework that inserts a Linear Residual Adapter (LRA) at each pruning site. Each LRA preserves the identity pathway of the original residual block and adds a lightweight affine residual correction. This correction is calibrated via closed-form least-squares regression on a small held-out set, without gradient computation, to approximate the missing residual update produced by the pruned block. Together with the preserved identity pathway, the resulting LRA output approximates the hidden state produced by the original block, thereby mitigating the distributional mismatch introduced by layer removal while avoiding the expensive attention and feed-forward computations of the removed blocks. The resulting LRAs support low-rank factorization and exact merging across consecutive pruned layers for additional compression, and combine naturally with parameter-efficient fine-tuning for further recovery beyond fine-tuning the pruned model alone. Experiments on five model families, six layer-selection criteria, and seven zero-shot benchmarks show that SHIFT-LLM consistently recovers accuracy lost to depth pruning across most configurations, achieving gains up to +15.7 points on Llama-3.1-8B-Instruct while requiring only a few hundred calibration samples and no gradient computation.

AI 服务暂时不可用，以下为来源摘要，待恢复后补全翻译：arXiv:2608.24935v1 Announce Type: new Abstract: Accurate identification of early-stage apple fruitlet anatomical structures, including the calyx, fruitlet body, and peduncle, is essential for robotic thinning, crop-load management, and other precision orchard operations. This study presents a lightweight multimodal vision-language framework that adapts TinyCLIP for fine-grained fruitlet anatomy classification in complex orchard environments. A dataset of 600 high-resolution RGB images collected from Scilate and Scifresh apple orchards was converted into 224 x 224 image patches and annotated for three anatomical classes. Domain-specific language prompts, such as ``a photo of a class,'' were used to guide multimodal alignment between orchard imagery and horticultural structures. A sliding-window inference strategy with a stride of 112 pixels aggregates patch-level predictions into spatial heatmaps, enabling interpretable whole-image localization of fruitlet components relevant to robotic thinning. Patch-level evaluation on an NVIDIA T4 GPU achieved F1-scores of 0.95 for calyx, 0.98 for fruitlet, and 0.85 for peduncle, with a macro-F1 score of 0.93. Deployment-oriented optimization using ONNX and TensorRT enabled efficient inference on NVIDIA Jetson hardware, preserved accuracy under INT8 quantization, and supported model sizes of approximately 127-137 MB with millisecond-level patch inference. These results demonstrate that lightweight vision-language models can provide interpretable and edge-deployable perception for automated fruitlet analysis and future robotic thinning systems. The source code and implementation details are publicly available at https://github.com/WilliamBu1/A-Lightweight-Vision-Language-Model-for-Early-Stage-Fruitlet-Classification-in-Apple-Orchards.

AI 服务暂时不可用，以下为来源摘要，待恢复后补全翻译：arXiv:2608.24945v1 Announce Type: new Abstract: Recent years have witnessed remarkable achievements of Large Language Models (LLMs) in multiple domains, while the excessive resource requirements of LLMs hinder the deployment on resource-constrained devices. Although model quantization stands out as an effective approach, conventional quantization approaches typically incur severe performance degradation due to uniform bit-width or simple heuristic sensitivity evaluation. In this paper, we propose a novel Fisher information-based Adaptive Mixed Precision Weight Quantization approach, i.e., FAMPWQ, which performs layer-adaptive weight quantization for effective LLM inference on commodity GPUs. First, we propose a system model with a novel Fisher information metric to measure the layer-wise sensitivity to quantization. Second, we propose a reinforcement learning-based bit-width allocator in FAMPWQ, which generates an adaptive bit-width allocation strategy based on the Fisher information sensitivity metric. Extensive experiments on 7 models and 5 benchmarks demonstrate that FAMPWQ significantly outperforms 7 baseline approaches in terms of PPL (up to 3.39 smaller), accuracy (up to 6.87% higher), and LLM-as-a-judge comparison (up to 76% win rate).

AI 服务暂时不可用，以下为来源摘要，待恢复后补全翻译：arXiv:2608.24938v1 Announce Type: new Abstract: Mixture-of-Experts (MoE) models scale capacity for strong quality while keeping per-token compute bounded through sparse expert activation. Yet low-latency MoE serving is increasingly challenging, because it spans two inference phases with fundamentally different bottlenecks: prefill is dominated by token-wise expert computation, whereas decode is constrained by memory traffic from the batch-wise activated expert set. However, existing training-free acceleration methods optimize only a single resource proxy, either the experts each token executes or the experts a batch activates, and either discard the excluded experts' contribution or leave it only implicitly approximated. In this paper, we propose ExFold, a unified training-free expert-folding framework for jointly accelerating MoE prefill and decode. ExFold casts both prefill and decode as one budgeted output-approximation problem: execute only a phase-specific constrained expert set while projecting the contribution of budget-excluded experts onto retained experts using calibrated scalar projectors. Motivated by the observation that many expert outputs are directionally aligned but differ in magnitude, ExFold calibrates a pairwise scalar-projector matrix on unlabeled data and uses it at inference time to fold excluded expert contributions into retained experts. Under this view, prefill acceleration becomes token-level Top-K folding, and decode acceleration becomes batch-level expert-pool folding. The two phases differ only in how retained experts are selected, while excluded contributions are recovered by one shared folding mechanism. We implement ExFold as a plug-and-play plugin in vLLM, with a lightweight expert-folding CUDA kernel, delivering up to 1.41x TTFT and 2.45x TPOT speedups while retaining about 99% of the original average quality.

AI 服务暂时不可用，以下为来源摘要，待恢复后补全翻译：arXiv:2608.24936v1 Announce Type: new Abstract: We present GreenLeaf Law Embed Tiny, a 0.6B parameter embedding model for legal domain retrieval. GreenLeaf-Tiny achieves 75.11% on the Massive Legal Embedding Benchmark (MLEB) and 64.38% on MTEB(Law, v1),demonstrating competitive performance among models under 1B parameters. Our approach combines a two-stage training pipeline that first distills knowledge from a larger teacher model into a compact student architecture, then applies domain-specific fine-tuning with hard negative mining; a carefully curated dataset of 3.4 million query-passage pairs, including 150,000 human-curated samples across diverse legal jurisdictions; and an efficient inference architecture supporting multiple quantization levels (BF16, INT8, binary) enabling deployment in resource-constrained environments. We provide detailed analysis of our training methodology, architectural choices, and comprehensive evaluation across legal retrieval tasks. Our results demonstrate that domain-specific training with high-quality data can improve performance for specialized domain applications

AI 服务暂时不可用，以下为来源摘要，待恢复后补全翻译：Z.ai has released GLM-5.3-Flash, the first natively multimodal model in the GLM-5 series — a 320B-total / 18B-active MoE with a 1,048,576-token context window, MIT-licensed weights on Hugging Face, and API pricing at $0.15/M input and $0.50/M output. It scores 84.3 on Terminal-Bench 2.1 and 63.4 on DeepSWE v1.1, using hybrid KDA linear plus NoPE sparse MLA attention to cut attention compute ~3× and KV cache 4.4× versus GLM-5.3. The post Z.ai Releases GLM-5.3-Flash: A 320B-A18B Natively Multimodal MoE With a 1M-Token Context appeared first on MarkTechPost.

AI 服务暂时不可用，以下为来源摘要，待恢复后补全翻译：Learn proven strategies to speed up your AI agent: reduce latency, optimize LLM calls, enable parallelism, and improve UX. Expert tips from LangChain.

AI 服务暂时不可用，以下为来源摘要，待恢复后补全翻译：arXiv:2608.23831v1 Announce Type: new Abstract: While reinforcement learning (RL) allows generalist robot policies to continually improve during deployment, the large model size of modern generalist policies, such as VLAs, poses a fundamental obstacle to effective RL improvement. In particular, their severe inference latency---which can lead to pauses or jerky movements---can alter the effective environment dynamics and, if not correctly accounted for, break the Markov assumption that RL relies on, causing standard RL algorithms to fail completely. In this work, we introduce a latency-aware framework, Asynchronous RL with Intermediate Information (ARLI), that enables RL-based improvement of generalist policies under inference delays. Our framework builds on asynchronous inference approaches, which interleave action generation with execution to hide latency, and addresses its incompatibility with RL by providing a low-latency RL policy design that maximizes reactivity within the inference window through two contributions: state augmentations that restore near-Markovian structure by incorporating committed actions and a mid-inference observation. We evaluate our approach across simulated and real-world manipulation tasks, and find that it enables effective finetuning under inference delays where standard RL fails entirely, even matching or exceeding the performance of standard RL in idealized no-latency settings.

AI 服务暂时不可用，以下为来源摘要，待恢复后补全翻译：arXiv:2608.23634v1 Announce Type: new Abstract: Many few-shot adaptation methods for vision-language models classify with a convex combination of the zero-shot text prototype and the mean of the K labelled image features, with a single blending ratio routinely tuned on held-out labels, often on the test set itself. We ask what the family's own bias-variance justification invites: what is the right ratio, can it be estimated without validation data, and is finding it where the performance is? First, the ratio minimising prototype mean-squared error has a closed form whose support-set plug-in is exactly a positive-part James-Stein coefficient shrinking towards the text prototype. Across 4,800 cells (ten datasets, five backbones including SigLIP, five shot counts, five seeds, four prompt tiers) this theoretically optimal ratio is a reliable estimate of the wrong quantity: on the 950 primary-tier cells where it is defined it trails a test-set-oracle ratio by 8.5 points. It saturates near 1, discarding the text prior for a nearest-class-mean classifier, because 78% of the text-image prototype distance it treats as bias is a class-independent offset that the arg max largely cancels. We prove the mechanism and bound its share of the damage at 26% by a counterfactual. Second, leave-one-out on the support set alone sets a ratio landing within 0.9 points of the oracle blend, so it is estimable without validation data. Third, validation-free linear probes beat even the oracle-tuned blend: CLAP by +1.9 points and LP++ by +1.5 on average, and at K >= 4 all four validation-free baselines sit above the oracle, the linear probes by margins excluding zero. These results locate the ceiling in the model class, not the hyperparameter: the ratio can be set near-optimally for free, and it is still not where the performance is. Code, cached features, per-cell records: https://huggingface.co/datasets/Liangzhi-Li/clipbench-blending

AI 服务暂时不可用，以下为来源摘要，待恢复后补全翻译：arXiv:2608.23794v1 Announce Type: new Abstract: Mixture-of-Experts (MoE) scales language models by routing each input through a small set of independently parameterized experts. We show that copying this design into convolutional networks fails for a structural reason: parallel convolutional experts that read the same input channels learn nearly identical filters. We therefore move the expert axis from operator duplication to channel selection. We introduce Mixture of Channel Experts (MoCE), a structured sparse channel-mixing layer, inspired by MoE, that replaces pointwise (1x1) channel-reduction projections. In MoCE, an expert is a single output channel with a learned sparse support of k << C input channels. The selected channels are combined by a softmax whose temperature is predicted per input, so each expert can move between mean-like and max-like aggregation. A residual expert summarizes the unselected channels, and a load-balancing loss keeps channel coverage complete. MoCE replaces a dense projection whose cost is quadratic in C with a mechanism whose relative cost scales as k/C, and the predicted savings hold in measured wall-clock time. Across ResNet backbones on ImageNet-1K and CIFAR-100, transfer learning, EfficientViT, and a strong modern training recipe, MoCE matches or exceeds dense baselines and prior channel-selection methods while reducing MACs by 16.7% and end-to-end latency.

AI 服务暂时不可用，以下为来源摘要，待恢复后补全翻译：Model cards report quality under server-class, full-precision conditions. Those numbers rarely predict how the same model behaves on a phone. This week, Liquid AI released Pipette. It is an open-source platform for benchmarking foundation models on edge devices, built in partnership with Artificial Analysis as an independent methodology validator. Pipette treats on-device behavior as a […] The post Liquid AI Open-Sources Pipette: A Reproducible Benchmarking Suite That Measures On-Device Models, Quantization, Runtime and Hardware Together appeared first on MarkTechPost.

AI 服务暂时不可用，以下为来源摘要，待恢复后补全翻译：OpenAI says its new AI chip, Jalapeño, completes tasks more efficiently and returns responses faster than other AI systems, according to a blog post published on Tuesday. During a briefing with reporters, OpenAI hardware vice president Richard Ho said Jalapeño offers the "best of both worlds" with lower latency and higher throughput, as AI systems typically "have to make a trade-off between the two." First introduced in June, Jalapeño is an Application-Specific Integrated Circuit (ASIC) made in partnership with Broadcom. It's designed for AI inference - the process of running a trained AI model to complete a task or deploy an agent. To mea … Read the full story at The Verge.

AI 服务暂时不可用，以下为来源摘要，待恢复后补全翻译：This article is brought to you by Ollobot. From about 2017, individuals began to truly connect with the initial wave of companion robots. These devices had personality, moved around, joked, and answered when you spoke to them. Most early companion robots, however, were still limited by simple voice-command interactions and narrow functionality. Once the novelty wore off, many ended up sitting unused on shelves. As some of those companies went out of business and turned off their servers, many owners likened it to losing a pet. What Ollobot describes as “gentle intelligence” is a useful way to think about where the serious work in this category is going. Not toward more powerful assistants, but toward more present ones. The problem companion robots were trying to solve Loneliness is not a niche issue. According to one study, nearly one out of three elderly adults resides alone, meaning they do not have daily companions. Research also shows that children whose parents have migrated for work, leaving them in the care of relatives, were 2.5 times more likely to experience loneliness than children whose parents remain with them. Among working adults living alone in urban environments, similar patterns of social isolation emerge, even if they are less visible. Over the years, technology has time and again attempted to solve this problem via video calls, smart speakers, and messaging apps without much success. Those tools are geared towards communication between people that already have relationships. They do not create presence. They schedule it. That is the gap that a new generation of AI companion robots is being engineered to fill. Today’s AI robots are different Today’s companion robots are not just cute and cuddly. They are designed with psychological research, clinical insight and long-term interaction models to be truly useful in real homes. Three fundamental shifts define the current generation: From reactive to proactive response. Older robots relied on you speaking to them, but modern robots monitor a room with cameras, microphones, and surroundings sensors to initiate interactions without your input, and they can pick up on your emotions. From function-oriented to emotion-oriented design. The original pitch for companion robots was about what they could do. The question driving the serious work now is how they make you feel, which is a harder engineering problem and a more honest framing of what the product is actually for. From standalone hardware to connected ecosystems. Leading brands are creating platforms rather than devices with software included as a built-in layer and remote access from the beginning. The global AI companion market size was valued at US $36.8 billion in 2025 and is projected to grow from $48 billion in 2026 to $318 billion by 2033, at a compound annual growth rate of 31 percent from 2026 to 2033. Three household scenarios and interaction models Ollobot’s advanced AI family companion robot OlloNi SS1 addresses a number of gaps in what existing technology offers. Elderly individuals living alone. The combination of proactive interaction, fall detection, and persistent presence addresses both safety and companionship without the social overhead of asking family members to check in more frequently. Children in households where parents work far from home. The SS1 functions as a consistent companion that already knows a child, their preferences, their moods, and their routines. The remote connection features allow parents to stay present without requiring a scheduled call, and the life recording system gives them a passive window into their child’s days that feels less clinical than a monitoring camera. Single professionals living alone in cities. The SS1 adapts to daily routines, builds up a preference model over time, and provides ambient social presence without demands. OlloNi SS1 adapts to daily routines over time.Ollobot What OlloNi SS1 is doing differently? Ollobot’s goal in building intelligent companion robots is to address the gaps in technology and capability, using innovation not to automate tasks but to fill emotional voids. Much of the robotics industry has historically pursued human imitation — machines that speak, look, or behave like people. The SS1 is instead designed around familiarity and long-term coexistence rather than realism. The system integrates multiple subsystems operating in parallel, including visual perception, audio processing, mobility control, and interaction management. It is equipped with a multi-chip AI 4K vision module capable of facial recognition and motion tracking. One small but revealing detail is the inclusion of a physical privacy cover for the camera — a mechanical solution to concerns that software settings alone may not fully resolve. OlloNi SS1 can actively integrate into family activities, and it can autonomously move closer to capture memorable moments or reposition itself to remain engaged in ongoing interactions.Ollobot The robot supports advanced mobility across multiple indoor surfaces, including wooden floors, ceramic tiles, and low-pile carpets, with slope climbing capability up to 3.5 degrees. Rather than remaining in a fixed location, it can move naturally throughout the home to stay close to household members as daily activities unfold. For example, the OlloNi SS1 may greet family members when they arrive home, follow an older adult from the living room to the kitchen while continuing a conversation, remind a child to take a study break after a prolonged period of inactivity, or notice that someone appears unusually quiet and gently check in. During family activities, it can autonomously move closer to capture memorable moments or reposition itself to remain engaged in ongoing interactions. The robot continues to evolve over time, with over-the-air updates that deliver new features, performance improvements, and AI enhancements It also incorporates fall detection with optimized accuracy for safety monitoring scenarios. A 6-microphone array enables omnidirectional voice pickup with an effective voice capture range of up to 5 meters, supporting reliable wake-word detection and far-field interaction. To support continuous companionship, much of the robot’s AI processing takes place directly on the device through its “heart module” architecture, with 16 GB of memory and 64 GB of local storage. This enables the system to retain household memories, recognize familiar faces, and respond with lower latency, making interactions feel more natural even during everyday routines. Because companion robots are expected to remain available throughout the day rather than only during brief interactions, the SS1 is designed for extended operation, offering up to 12 hours of standby time and around 5 hours of active interaction on a single charge. This allows it to accompany users through meals, conversations, playtime, and other daily activities without frequent interruptions. To support engaging interactions, much of the robot’s AI processing takes place directly on the device through its “heart module” architecture.Ollobot Like the relationships it is designed to build, the robot continues to evolve over time. Running on Android OS with over-the-air (OTA) updates, the system continuously receives new features, performance improvements, and AI enhancements, allowing its capabilities to grow alongside the household it serves. The robot’s behavioral model also improves over time. Rather than reacting to isolated commands, it attempts to establish a baseline understanding of household routines and individuals. Changes in behavior — prolonged quietness, unusual inactivity, or emotional cues — become triggers for interaction. Presence instead of utility Several features in the OlloNi SS1 illustrate this emphasis on presence and continuity in its interactions. The system can identify different household members, including pets, and adapt responses accordingly. Remote communication features allow family members to connect through the device without treating every interaction like a scheduled call. Environmental sensors support contextual reminders tied to weather or room conditions. Its “2+1” multi-display configuration is also designed around emotional communication. Two circular side displays function as expressive “emotional eyes,” while a separate primary display handles information and structured interaction. The separation allows emotional signaling and functional communication to operate independently, creating more intuitive nonverbal interaction even when no dialogue is taking place. The robot’s behavioral model improves over time. Rather than reacting to isolated commands, it attempts to establish a baseline understanding of household routines and individuals.Ollobot The SS1 also includes an automated life-recording system built on facial recognition and behavioral-event detection that can capture moments such as laughter, physical closeness, or group interaction automatically. An integrated AI vlog engine can then organize those moments into edited short-form videos with automated sequencing and soundtrack generation. The design intent is to preserve spontaneous domestic moments without requiring active documentation behavior from users. An integrated AI vlog engine can organize recorded moments into edited short-form videos with automated sequencing and soundtrack generation Visual data is processed primarily on the device through the SS1’s on-device AI architecture, with household memories stored locally and managed within Ollobot’s proprietary ecosystem instead of being shared with third-party smart home platforms. Access to recordings and live feeds is restricted to authorized users through the companion app, while encrypted communication helps protect data during remote access. Users also retain direct control over recording preferences, and the physical camera privacy cover provides an additional hardware-level safeguard whenever visual monitoring is not desired. Learn more at ollobot.com. Remote communication is similarly structured around persistence rather than transaction. Traditional video calls are episodic and screen-bound; the SS1 instead acts as a continuously present interface embedded inside the household environment. Through autonomous mobility, environmental awareness, and persistent household memory, remote family members interact with an ongoing domestic context. The larger shift to “gentle intelligence” Ultimately, gentle intelligence is not about making robots behave more like humans — it is about helping them fit more naturally into human lives. Each OlloNi SS1 unit develops a unique behavioral profile based on its household. Two units running in different homes for a year will have become meaningfully different from each other, shaped by the specific people, habits, and rhythms of where they live. That kind of long-term personalization is what early companion robots never had. It is also what makes the difference between a product that ends up on a shelf and one that actually earns its place in a home. Learn more at ollobot.com.

AI 服务暂时不可用，以下为来源摘要，待恢复后补全翻译：Jalapeño is a custom inference chip from OpenAI that delivers faster, more power-efficient AI inference, with higher throughput and lower latency for modern models.

AI 服务暂时不可用，以下为来源摘要，待恢复后补全翻译：arXiv:2608.21393v1 Announce Type: new Abstract: Running large language models inside enterprise environments has always bumped up against a practical wall: the data lives in one place, the AI horsepower sits somewhere else, and moving sensitive records between the two creates real headaches around latency, security, and regulatory exposure. IBM's Spyre accelerator PCIe inference card built for LinuxONE and the broader IBM Z family changes that equation. In this paper we lay out a six-subsystem RAG architecture that runs entirely on IBM LinuxONE, using Spyre for generative inference, the Telum II on-chip accelerator for lightweight classification tasks, and Red Hat OpenShift for container orchestration. Every piece of the pipeline from query intake through vector retrieval, prompt assembly, LLM inference, compliance filtering, and response delivery stays within a single LinuxONE system, so sensitive data never has to leave the hardware perimeter. We walk through the design choices behind each subsystem, dig into the Spyre compilation and serving stack, explain how LinuxONE's Secure Execution technology extends confidential-computing guarantees to AI workloads, and benchmark the architecture against cloud-GPU and on-premises alternatives. Early analysis points to end-to-end RAG latencies under two seconds and up to a 20x reduction compared to off-platform inference, all while keeping the strong encryption and auditability posture that regulated industries actually need.

AI 服务暂时不可用，以下为来源摘要，待恢复后补全翻译：arXiv:2608.21375v1 Announce Type: new Abstract: Heterogeneous agentic retrieval-augmented generation (RAG) systems increasingly orchestrate external APIs, internal databases, vector stores, and graph stores. Exposing all tool descriptions to an LLM agent, or selecting tools only by vector similarity, causes two costly failures: over-fetching, which increases payload size, token use, and latency, and under-fetching, which omits fields needed to answer the query. We present SchemaRouter, a lightweight routing layer that represents tools, endpoints, parameters, response fields, domain concepts, units, provenance, and license policies as a schema graph. Given a query, SchemaRouter emits an executable tool plan specifying which tools to call and which fields to retrieve. A small LLM extracts intent, concepts, and source constraints, while field selection is deterministic over the graph through intent-group projection and concept-field matching with an alias layer. On a materials-science benchmark of 110 queries, SchemaRouter achieves answer accuracy of 0.71, matching fetch-everything within overlapping confidence intervals and exceeding prompt-all's 0.66, though their intervals overlap. It uses 227 retrieved-context tokens versus 2,066 for fetch-everything and achieves 2.7x lower end-to-end latency than prompt-all. It also obtains the best tool-exact rate of 0.93 and parameter validity of 1.0. SchemaRouter grounds provenance and license information in 62 percent of answers, compared with approximately 0 percent for all baselines. We also find that minimizing selected-field count is counterproductive: it reduces answer accuracy to 0.56 with negligible token savings, while recall-preserving projection restores top accuracy. SchemaRouter improves efficiency, schema-size-independent scaling, and verifiable provenance/license-grounded answering at competitive accuracy.

AI 服务暂时不可用，以下为来源摘要，待恢复后补全翻译：arXiv:2608.21362v1 Announce Type: new Abstract: Transformer-based large language models (LLMs) incur high prefill latency because key-value (KV) tensors must be recomputed for each request. Existing prefix-caching systems reduce this cost but require prompts to share a leading contiguous prefix, limiting effectiveness when shared content appears at arbitrary positions. We present KVBoost, a chunk-level KV cache reuse system for HuggingFace-compatible decoder models that enables reuse regardless of content position. KVBoost introduces a dual-hash keying scheme that separates positional identity (prefix hash) from content identity (content hash), supporting both exact and approximate cache matches. To address attention boundary errors from independently cached chunks, KVBoost employs two repair strategies: SelectiveRecompute, which re-encodes boundary regions, and CacheBlendRecompute, which identifies and recomputes high-deviation tokens after a probe pass. The system further incorporates asymmetric KV quantization (int8/int4), adaptive chunk boundary splitting, and importance-weighted eviction under a fixed memory budget. Evaluated on Qwen/Qwen2.5-3B over 1,000 bug-localization samples, KVBoost achieves a 4.49x reduction in time-to-first-token (142.4 ms vs.\ 639.1 ms) and outperforms prefix caching by 16%, with no loss in accuracy (99.2% vs.\ 99.1%). KVBoost provides a practical, memory-bounded inference acceleration layer compatible with RoPE-based models without architectural modification.

AI 服务暂时不可用，以下为来源摘要，待恢复后补全翻译：Unified multimodal models that understand, reason over, and generate interleaved text–image sequences remain structurally fragmented: existing approaches either sacrifice visual fidelity through discrete tokenization, impose structural asymmetry by combining causal text generation with iterative diffusion-based denoising, or degrade pretrained understanding when adapting vision-language models for generation. We observe that autoregressive normalizing flows are autoregressive Transformers—sharing the same causal mask, KV-cache mechanism, and left-to-right structure as LLMs—making them the most…

AI 服务暂时不可用，以下为来源摘要，待恢复后补全翻译：Physical artificial intelligence is emerging as the next major phase of AI. These systems not only generate content or analyze data but also perceive, reason about and act in the physical world. The opportunity is massive, but so are the operational, data, latency and lifecycle-management challenges. That is why Amazon Web Services Inc. last month […] The post Physical AI’s moment has arrived – but moving from demo to deployment is the hard part. AWS wants to fix that appeared first on SiliconANGLE.

AI 服务暂时不可用，以下为来源摘要，待恢复后补全翻译：Moonshot AI’s Kimi K3 is a 2.8-trillion-parameter open-weight model built with a Mixture-of-Experts architecture. It activates only a small fraction of its parameters per token, helping reduce inference costs while delivering strong coding and agentic performance. K3 combines near-frontier capabilities, open weights, and lower API pricing, making it an interesting alternative to proprietary models. In […] The post How to Use Kimi K3: Moonshot AI’s 2.8T Open-Weight Model appeared first on Analytics Vidhya.

AI 服务暂时不可用，以下为来源摘要，待恢复后补全翻译：arXiv:2608.20515v1 Announce Type: new Abstract: Generative video compression can recover rich visual details at low bitrates, but simultaneously achieving high temporal consistency and low inference cost remains challenging. To address this issue, we propose DiffVC-ONE, a diffusion-based generative video compression framework built on a one-step Video Diffusion Transformer. First, we introduce a Unified Unidirectional Latent Compressor that uses a shared model to efficiently and uniformly compress compact latent slices. We then develop a Video DiT-based One-Step Diffusion Enhancer that uses the reconstructed latent slices as content anchors and performs single-step spatio-temporal perceptual enhancement over an entire group of pictures. Finally, a Hybrid Condition Generator extracts structural, strength, and semantic conditions from the reconstructed content and quantization information. These conditions preserve faithful regions, control the degree of generative enhancement, and supplement content-aware perceptual details during one-step diffusion enhancement. Extensive experiments on multiple standard benchmarks demonstrate that DiffVC-ONE achieves state-of-the-art perceptual quality and temporal consistency with low inference cost.

AI 服务暂时不可用，以下为来源摘要，待恢复后补全翻译：arXiv:2608.20350v1 Announce Type: new Abstract: Traditional industrial agents rely on modular pipelines, including Router, Retriever, Planner, Executor, Responder, Reviewer, and other components. These systems often fracture into a labyrinth of ad-hoc patches, leading to cascading errors and high latency. We propose OneModel, an applicable paradigm shift from external workflows to internalized knowledge representation. Unlike modular systems that slice fluid user intents into static steps, OneModel consolidates complex business logic and SOPs directly into the model parameters. Through Continual Pre-training (CPT) and logic-compilation SFT, we transform fragmented business rules into intuitive model reasoning within a unified attention space. Deployed in our global financial service system, OneModel effectively breaks the trade-off between latency, accuracy, and complexity. Online A/B testing demonstrates an end-to-end latency reduction of more than 50 percent, from 18.7 seconds to 8.0 seconds, while the Intelligent Resolution Rate (IRR) increases from 64.3 percent to 83.3 percent. The results show that OneModel can replace brittle engineering logic with internalized cognitive intuition, offering a scalable blueprint for transitioning industrial agents from complex, error-prone workflows to unified model architectures.

AI 服务暂时不可用，以下为来源摘要，待恢复后补全翻译：arXiv:2608.20439v1 Announce Type: new Abstract: Neural PDE operators are increasingly trained on reusable solver archives, yet validation often relies on clean prediction error and parameter-agnostic plausibility checks. We introduce cross-parameter relinking, a data-poisoning primitive that makes a triggered input select a valid solution from the same PDE family under an incorrect physical parameter. We term this a wrong-physics backdoor: the output remains physically plausible but is wrong for the intended parameter. The attack exploits tensor-to-parameter provenance failures in multi-parameter archives by stamping the surrogate input and relinking its supervision to a cached alternate-parameter solution for the same latent sample. Across 476 attack campaigns, we evaluate Burgers, advection-diffusion, two-dimensional Navier-Stokes, and an elliptic Poisson case. Fourier Neural Operators and DeepONet provide the primary evidence, with Transformer, GRU, and LSTM models as support. FNO reaches a backdoor success rate of 1.0000 on both advection-diffusion and two-dimensional Navier-Stokes while retaining low clean relative L2 error. Clean-label, label-only, and shuffled controls show that high attack success alone is insufficient: successful attacks must move predictions toward the intended alternate-physics target while preserving bounded clean error. These results expose a structural validation gap: smoothness or generic solver-like behavior is insufficient unless the provenance of the intended physical parameter is also verified.

AI 服务暂时不可用，以下为来源摘要，待恢复后补全翻译：arXiv:2608.20427v1 Announce Type: new Abstract: Dense causal attention remains expensive at long context even when implemented with highly optimized exact kernels. We study BF1, a deterministic block-aligned dyadic sparse-attention route that combines a small exact local neighborhood, a global first block, and logarithmically spaced historical blocks. The route is related to prior log-sparse and dilated attention patterns; our contribution is a correctness-gated pretrained-model retrofit, a matched topology-control study, and a systems characterization that connects per-layer sparsity to whole-model latency. For fixed block width, every converted layer uses O(n log n) selected token interactions and has O(log n) graph communication depth. On an NVIDIA RTX PRO 6000 Blackwell GPU, an optimized BF16 implementation crosses dense attention between 2K and 4K tokens and reaches a 10.91x per-layer prefill speedup at 32K. Retrofitting eight of 28 Qwen3-0.6B attention layers lowers warm whole-model time to first token by 7.7%, 11.3%, and 15.3% at 8K, 16K, and 32K, respectively, while the remaining dense layers keep the complete model asymptotically quadratic. Under a matched 1,000-step, 16.384M-token adaptation protocol, BF1 ranks first across three training seeds: mean report perplexity is 1.68639 versus 1.69154 for a matched static-random nonlocal graph, 1.69258 for dense continued training, and 1.81505 for equal-budget local sliding. At seed 1234, the packed-report paired interval places Dense-CT 0.3169-0.4055% above BF1 and static-random graph 17 0.2441-0.3642% above BF1. These results establish BF1 as a reproducible sparse operator and selective retrofit primitive with real long-context systems value. This paper evaluates numerical correctness, selected-interaction scaling, kernel performance, partial-model inference, and matched next-token language modeling.

AI 服务暂时不可用，以下为来源摘要，待恢复后补全翻译：arXiv:2608.20397v1 Announce Type: new Abstract: Agentic large language models (LLMs) on the Model Context Protocol (MCP) re-encode verbose tool schemas every turn, so prefill - quadratic in sequence length - dominates time-to-first-token (TTFT) as the tool registry grows. Nexus's primary lever is to decouple routing from the schema-prefill cost: an INT8 semantic lookaside buffer (SLB) with a calibrated cross-encoder margin gate selects tools by retrieval, and arguments are generated over a compressed textual signature (median 19 tokens) rather than over spliced key/value (KV) cache. This path is depth-independent: routing accuracy stays near 89% as the registry scales to 250 tools - where a concatenate-all-schemas baseline overflows the context window entirely - and it reaches a first-argument token 1.66x sooner than a full-schema re-prefill at a ~80% main-context token saving. As a secondary, bounded lever we transplant a compiled schema KV block directly into the live context. This is fundamentally limited by rotary position embedding (RoPE) phase drift: an anchored splice is output-exact, but off-anchor placement corrupts attention, so beyond a threshold P=256 Nexus repairs the seam with a depth-adaptive suffix redecode that escalates to a full re-prefill. The resulting never-regress property is a guarantee on output fidelity (top-1 agreement, D_KL approx. 0) - not on latency, which can dip to 0.98x before converging to parity - alongside a 1.1-1.7x TTFT speedup at moderate depth that narrows to parity at deep context. Two negative results bound the design: the off-anchor RoPE fidelity boundary, and the failure of a reference-free drift gate to predict drift (Spearman rho = 0.193). All measurements are from one model tuple (Qwen2.5-14B-Instruct Q4_K_M) on Apple-silicon unified memory; the qualitative boundaries generalize, while the quantitative envelope is tuple-specific.

AI 服务暂时不可用，以下为来源摘要，待恢复后补全翻译：arXiv:2608.20379v1 Announce Type: new Abstract: Advances in large language models (LLMs) have fueled a wave of research into agency: the ability to reason, plan, and act. This effort has produced agentic frameworks that orchestrate perception, memory, and decision-making around powerful LLM backbones. With the advent of large multimodal models (LMMs), these systems can process and integrate diverse modalities, including images, audio, and video, thereby improving their real-world applicability. Yet, while surveys of LLM-based agents exist, the role of multimodality in shaping agency has not been systematically examined in recent years. This survey fills the gap by analyzing the impact of multimodality across the core functional modules of the agentic framework: perception, reasoning, planning, memory, and action. Using this lens, we trace the evolution from text-centric agents to multimodal frameworks, examine how modalities are integrated through delegated, late-fusion, and early-fusion architectures, and assess the emergence of agentic behaviors enabled by grounded perception and multimodal reasoning. We organize existing work through a modality-centric taxonomy that links architectural design choices to agent capabilities. Moreover, we review multimodal agentic systems across various application domains, including Robotics, GUI & Web Navigation, Multimedia Content Generation & Editing, and Long-form Video Understanding & Retrieval. Beyond capabilities, we analyze performance across these settings and discuss efficiency-scalability trade-offs, including training and inference costs, latency, and deployment constraints. By focusing on the impact of multimodality in agentic design, we aim to identify key gaps and chart a roadmap toward robust and general-purpose intelligent systems.

AI 服务暂时不可用，以下为来源摘要，待恢复后补全翻译：FreeToken splits MoE cache misses between PCIe fills and CPU execution using measured bandwidths, unlocking frontier models locally The post Meet FreeToken: An Edge-Native MoE Serving Engine that Runs 753B GLM-5.2 on a Single Workstation GPU appeared first on MarkTechPost.

AI 服务暂时不可用，以下为来源摘要，待恢复后补全翻译：Modern LLMs rely on quantization, pruning, distillation, and faster attention kernels, but production performance often depends most on KV cache management. As context windows grow, the cache consumes significant GPU memory, limiting concurrency, throughput, and latency. Two breakthroughs transformed this challenge: PagedAttention improves memory allocation, while RadixAttention enables efficient prefix reuse. Together, these techniques make […] The post PagedAttention vs. RadixAttention: Optimizing LLM KV Cache Management appeared first on Analytics Vidhya.

AI 服务暂时不可用，以下为来源摘要，待恢复后补全翻译：arXiv:2608.19553v1 Announce Type: new Abstract: Vision--language models can identify the correct referent while returning an imprecise bounding box. We study whether a frozen direct-answer model can use its own prediction to allocate one additional localized observation without accessing target annotations at inference. Label-free precision refinement (LFPR) routes predicted-small regions to a higher-resolution pass, re-grounds the expression inside a context crop, admits a candidate only under fixed geometric guards, and returns a fixed coordinate-wise midpoint. We report results across three evidence tiers. On 31,921 retrospective Ref-L4 expressions, LFPR raises mAcc$_{0.5:0.95}$ from 72.947\% to 76.013\% ([email protected] 88.531\%$\to$89.725\%, [email protected] 55.788\%$\to$61.142\%). A frozen transfer to 30,969 RefCOCO/RefCOCO+/RefCOCOg expressions improves every dataset at [email protected], mAcc, and mean IoU (pooled mAcc $+0.645$, [email protected] $+0.817$), while [email protected] is unchanged overall: routing alone gains $+1.162$ points there, but crop, guards, and fusion give back $-1.192$, offsetting rather than showing no strict-IoU effect. A prospective, image-disjoint Flickr30K Entities evaluation improves every endpoint (mAcc $+0.973$, [email protected] $+1.022$), more strongly under a single-box variant (mAcc $+2.575$, [email protected] $+3.689$). The same operator applied to two released grounding specialists improves every endpoint ([email protected] $+1.569$/$+6.716$ for EGM-4B/8B) at roughly twice the latency, composing with specialist training rather than replacing it. A genuine unguarded control (guard removed from the same candidates) underperforms the incumbent on every metric, showing the guard is load-bearing. Together, these results show that referent selection and boundary precision are partially separable, with different components moving opposing regions of the IoU curve -- behavior a single threshold cannot reveal.

AI 服务暂时不可用，以下为来源摘要，待恢复后补全翻译：arXiv:2608.19535v1 Announce Type: new Abstract: Retrieval-augmented generation (RAG) improves language-model responses by grounding generation in external passages, which comes with overhead: retrieved context lengthens the prompt, increasing prefill work, KV-cache footprint, memory traffic, latency, and energy. Context compression offers a natural remedy by pruning retrieved text before generation. However, state-of-the-art context-compression methods are typically used with a fixed compression budget, or with the rate selected offline and then applied at inference time. This static view ignores both workload variation and the live state of the edge device. On an edge SoC, compression is not free: the compressor itself runs on the same SoC and consumes latency and energy that can offset any generation savings. This paper proposes a vision for telemetry-informed adaptive compression in edge RAG, grounded in experimental evidence. We characterize the compression tradeoff on the NVIDIA Jetson AGX Thor using Llama and Qwen generators, Natural Questions and HotpotQA datasets, and LLMLingua-2 compression. Our measurements show that generation dominates the RAG budget for larger models, reaching roughly 90% of per-query latency and 91% of GPU energy for 7B-8B generators. Exploring the impact of the compression rate reveals an adaptive operating region: mild compression can miss energy opportunities, and overly aggressive compression can hurt inference quality. Intermediate compression can reduce GPU energy by up to 53.2%, and SoC energy by up to 48.2%, with negligible quality loss. We argue for runtime policies that dynamically manage compression, guided by workload features and edge telemetry.

AI 服务暂时不可用，以下为来源摘要，待恢复后补全翻译：Amazon Bedrock now offers OpenAI GPT-5.6 models (Sol, Terra, and Luna) in more than 25 AWS Regions with cross-Region inference. Learn how US geographic and global inference profiles route requests for higher throughput, how to call the models with the OpenAI and Converse APIs, and how to configure IAM, quotas, and monitoring.

AI 服务暂时不可用，以下为来源摘要，待恢复后补全翻译：arXiv:2608.18147v1 Announce Type: new Abstract: Adaptive stochastic quantization (ASQ) is a recently introduced quantization approach that optimizes the Mean Squared Error (MSE) for a given input while preserving unbiasedness. It is designed to alleviate the communication and memory bottlenecks of modern data and machine learning workloads, including model, gradient, and KV-cache compression and nearest-neighbor search. Further, practical systems can then compress quantized data with a lossless entropy encoder. However, existing unbiased methods, including ASQ, choose their quantization values without considering this later encoding stage, leaving accuracy on the table. We formulate the Entropy Constrained Adaptive Stochastic Quantization (ECASQ) problem, which jointly selects adaptive quantization values to minimize MSE under an entropy budget and an unbiasedness constraint. We give an optimal dynamic program with $O(sd^2)$ time and $O(d^2)$ space for a length-d vector and at most s quantization values, and a GPU-friendly approximate dynamic program with $O(sd^2)$ time and $O(d)$ space. The approximation guarantees that the solution has an MSE no larger than the optimal solution that uses one fewer bit of entropy per entry. We also provide an iterative refinement procedure for the approximation solution that, in our experiments, yields near-optimal results while retaining a substantial speed advantage over our solver for the optimal solution.

AI 服务暂时不可用，以下为来源摘要，待恢复后补全翻译：Before you change a setting, decide what better retrieval means for your workload. The right document at rank one, more candidates for a reranker, lower latency, and a smaller memory footprint each favor different settings, so pick your goal first. If your labeled queries can’t detect the improvement you’re chasing, you won’t be able to tell whether a change helped. Some settings are there to verify correctness, not to tune performance. If a vector is unindexed, a sparse vector is missing the IDF modifier, or the BM25 average length is wrong, the results are invalid. Any benchmark or comparison you run after that will reflect a broken setup. This article shows you how to check each setting and what the correct state looks like.

AI 服务暂时不可用，以下为来源摘要，待恢复后补全翻译：AI inference infrastructure is becoming a system-level challenge as organizations move generative and agentic applications into production. Graphics processing unit performance remains essential, but storage latency, network bandwidth, data movement and power consumption increasingly determine the cost and speed of producing tokens. The requirements also vary by workload. Interactive chat prioritizes latency, batch inference emphasizes […] The post The AI inference race moves beyond GPUs to reshape data center infrastructure appeared first on SiliconANGLE.

AI 服务暂时不可用，以下为来源摘要，待恢复后补全翻译：arXiv:2608.16978v1 Announce Type: new Abstract: Turning a frontier vision-language model into a robot policy usually means fine-tuning it to emit an action representation it never saw in pretraining, which throws away much of the reasoning that made the model worth reaching for. We go the other way and keep the VLM frozen. It writes the policy as a short Python control function, with no demonstrations and no fine-tuning. Writing that code once is open-loop, though. Existing closed-loop methods react at the wrong level: they retry a fixed policy or pick a different subtask, but never rewrite the code that failed. VLCP closes the loop where the failure actually lives, on the control code, within a single episode. Every $K$ steps the VLM re-observes the scene from multi-view RGB, proprioceptive state, and a state delta, then rewrites the control function from what it just saw, so a failure is caught before it compounds. We evaluate on a 57-task MuJoCo/RoboVerse sweep. This training-free policy reaches $35.1\%$ pooled success, against $3.5\%$ for the identical system queried once per episode. That tenfold gap holds with non-overlapping confidence intervals in every scene family. The gain traces to a $27.3\%$ within-episode recovery rate on failed grasps: a miss an open-loop controller would carry to the end of the episode gets re-observed and fixed at the next replan. And the loop stays cheap. A median $84\%$ of input tokens hit cache, an episode needs only about $10$ compact queries, and control blocks written during any replan persist to a cross-episode skill library reused in later prompts.

AI 服务暂时不可用，以下为来源摘要，待恢复后补全翻译：arXiv:2608.17050v1 Announce Type: new Abstract: Methods for improving knowledge use in large language models typically fall into two regimes. Non-parametric retrieval offers flexible access to external knowledge, but adds retrieval latency, context overhead, and only shallow integration with the backbone. Parametric adaptation is efficient at inference time, but entangles knowledge with model weights and can be hard to update, audit, or transfer. Engram-style hashed memory occupies a middle regime: it stores learned information in an external, addressable table, yet consumes that table through a small learned reader. This raises a basic question: when such a memory is moved across backbones, what matters more, the frozen memory itself or the target-side reader? We study this question through cross-model frozen-memory extraction, in which a memory trained on a source model is frozen and attached to a different target model, with only a lightweight reader trained. Ablations show that learned memory content and correct addressing both matter, but the transferred table becomes useful only through a reader aligned to the target model. In downstream question answering tasks, a dual-layer, four-branch reader nearly closes the gap between same-model and cross-model reuse, achieving an average score of 38.8 under our controlled evaluation protocol. Moreover, when the provider reader is directly compatible with the target interface, the frozen artifact can provide substantial utility without target-side training, while optional reader adaptation yields further improvement. These results suggest that Engram can serve as a reusable external knowledge artifact, provided that the target has access to a compatible reader interface; target-side adaptation can further improve alignment when direct reader reuse is insufficient.

AI 服务暂时不可用，以下为来源摘要，待恢复后补全翻译：arXiv:2608.17071v1 Announce Type: new Abstract: We present KernelArc, a multi-agent framework for autonomous GPU kernel optimization across heterogeneous workloads. Strategy-specialized agents run in parallel and coordinate through conclusions-only shared memory, a deterministic benchmark guard, and read-only cross-agent state with plateau-triggered drafting. We evaluate \kernelarc{} on NVIDIA H100 and B200 GPUs using category-representative SOL-ExecBench workloads. The resulting implementations span custom BF16 GEMM, static cuBLASLt Expert-API configuration tables, fused mixture-of-experts backward, shape-gated decoder-layer fusion, native NVFP4 grouped-query attention, and paged prefill attention. At the public SOL-ExecBench leaderboard snapshot recorded on July~30, 2026, these submissions ranked first on representative L1, L2, Quantization, and FlashInfer tasks. The trajectories support the paper's central motivation: shared multi-agent search can broaden exploration and reach stronger incumbents within a fixed candidate budget, while the value of individual coordination features depends on the kernel and optimization stage.

AI 服务暂时不可用，以下为来源摘要，待恢复后补全翻译：GPU stress testing and diagnostics Pantheon tests GPU compute, memory, cache, interconnect, and power behavior. Run focused workloads, capture telemetry, and keep the results for comparison. Download Pantheon View bench…

AI 服务暂时不可用，以下为来源摘要，待恢复后补全翻译：arXiv:2608.14604v1 Announce Type: new Abstract: Small language models in the ten to one hundred million parameter range are attractive for on device inference, rapid experimentation, and controlled scientific study, yet most of them reuse the standard transformer block without adaptation to the small scale regime. We present Wiola, a decoder only language model whose novelty is concentrated in three drop in components of every layer. First, Spiral Rotary Positional Encoding perturbs the standard rotary frequencies by a slowly growing per dimension factor so that phase trajectories fan outward, improving long range discrimination while adding no parameters. Second, Gated Spiral Attention introduces a per head, content adaptive scalar gate derived from a causal cumulative statistic of the query stream, providing an implicit and differentiable form of soft head selection at negligible cost. Third, the Butterfly feed forward block replaces the conventional expansion layer with a multiplicative interaction and an intra block bypass path, matching the parameter count of a four times gated linear unit block while improving gradient flow in shallow stacks. We formalize each component, derive exact parameter and computation budgets, and prove that the gated attention admits an exact and numerically verified equivalence between full sequence training and cached autoregressive decoding, so that no approximation is introduced at inference time. We also describe a fully reproducible training and evaluation protocol on a standard tiny story corpus. The reference implementation is released as an open source package with weights ready publishing support.

AI 服务暂时不可用，以下为来源摘要，待恢复后补全翻译：arXiv:2608.14551v1 Announce Type: new Abstract: Large language models (LLMs) are increasingly used for title-abstract screening in systematic reviews, but their decisions lack calibrated uncertainty. We show that an auxiliary BERT+GCN classifier supplies a structured uncertainty signal that improves LLM screening efficiency, and we identify the prompt-delivery strategy that maximises the benefit-to-cost ratio. We evaluate five LLM prompt-delivery conditions on eight drug-class datasets from the Cohen (2006) benchmark using 3 seeds x 5-fold stratified cross-validation (600 fold-level results). A BERT+GCN model trained per fold classifies each test paper as INCLUDE, EXCLUDE, or MAYBE via two spectral tests (algebraic radical and categorical paradox). Conditions vary information content (none / label / full scores), selectivity (all papers vs. MAYBE only), and timing (proactive vs. reactive two-pass). A cross-model pilot against gpt-4.1-mini on three datasets tests cross-generation transfer. Three findings: (i) Full-context delivery yields significant gains in F1 (+0.011, paired Wilcoxon p=0.008) and WSS@95 (+0.050, p=0.039) at a 1.28x token-cost premium, while preserving recall. (ii) MAYBE-only routing is Pareto-optimal: highest mean recall (0.92) and AUC-ROC (0.54) at only 1.05x baseline cost -- one sixth of full-context overhead. (iii) The two-pass design escalates 22.2% +/- 8.8% of records yet never revises its decision (0% flip rate across all datasets and folds), giving decisive evidence that current instruction-tuned LLMs cannot self-triage. The cross-model pilot shows an identical +0.8% recall uplift for both LLM generations. A per-paper ablation across 20,796 observations shows the dual paradox test reduces empirically to a one-line logit-gap criterion. We release the full pipeline; the 600-run experiment replays in under one hour from cached LLM responses.

AI 服务暂时不可用，以下为来源摘要，待恢复后补全翻译：arXiv:2608.14614v1 Announce Type: new Abstract: As AI datacenters retire functional GPUs, vast quantities of still capable accelerators enter secondary markets. This paper investigates whether these retired GPUs can find a productive afterlife to form a DumpsterCluster that can serve modern LLM inference, and under what conditions such repurposing is economically viable and environmentally sustainable. We physically built a 128-GPU DumpsterCluster from scratch using only second-hand components and ran it for one year. At current market prices (\$22K for the DumpsterCluster vs. \$600K for an 8-GPU B200 system), the economic advantages are substantial. Through pipeline-parallel optimizations, our V100 based DumpsterCluster achieves competitive LLaMA-70B throughput, validating production viability. However, our deployment reveals critical context dependencies. Older GPUs consume significantly more energy per token, making total cost of ownership favorable only in regions with inexpensive electricity. Under grid-average carbon intensity, second-hand systems can produce approximately 4x higher total carbon emissions per token for 8B models, and over 40x for 70B models, compared to current-generation hardware. These findings show that GPU afterlife is not universally sustainable - hardware repurposing must be strategically coupled with low carbon energy sources. When deployed in regions with favourable energy economics and clean electricity, second-hand GPUs offer a viable pathway for expanding AI capacity while advancing affordability, energy security, and environmental responsibility.

AI 服务暂时不可用，以下为来源摘要，待恢复后补全翻译：arXiv:2608.14563v1 Announce Type: new Abstract: Forward-Pass-Only MLP training (FPO) adapts large language models without a backward pass through the model body, achieving 2.7--3.2x the throughput of standard fine-tuning at ~40% less peak training memory, while leaving off-domain benchmarks within seed-noise of baseline, a property that full-network fine-tuning does not reliably reproduce. FPO rests on a single empirical observation: at late layers of a transformer, the output-layer prediction error approximates the true gradient with cosine similarity 0.47--0.59 across six public models we survey. We introduce a two-minute diagnostic that quantifies this approximation per layer for any model, identifying where late-layer adaptation is viable. Informed by the diagnostic, FPO computes a single error signal at the output and applies it to each target layer. No signal is propagated between layers, and no autograd graph is constructed at any point. We evaluate FPO on three model families (OLMo-2-7B, Qwen3-8B, Falcon3-7B). Across all three, FPO produces in-domain perplexity improvement and leaves MMLU, ARC-Challenge, HellaSwag, and Winogrande within seed-noise of baseline. Localizing SFT to FPO's target layers to enter this regime is also feasible, but at 2.2x the wall-clock cost of FPO.

AI 服务暂时不可用，以下为来源摘要，待恢复后补全翻译：NVIDIA Nemotron 3.5 Lightning, an open model built for high-volume agentic workloads, is now available in Amazon SageMaker JumpStart. This post shows how to deploy the 30B Mixture-of-Experts model (3B active), which delivers up to 4x higher throughput and up to 30% faster task completion for always-on agents.

AI 服务暂时不可用，以下为来源摘要，待恢复后补全翻译：The latest stable kernel also brings filesystem and I/O improvements, and substantial new support across AMD, Intel, Apple, Nvidia, USB4, and laptop hardware.