Knowledge catalogue
arXiv:2608.27205v1 Announce Type: new Abstract: Nitrogen rates for winter wheat are set before the season, under unknown prices and weather. The standard UK advice does not respond to prices, yet rece
arXiv:2608.27370v1 Announce Type: new Abstract: Language model pretraining has become almost synonymous with prohibitive cost, placing it out of reach for much of the academic and open-source communit
arXiv:2608.27302v1 Announce Type: new Abstract: Accurate identification of the correct view or angle in cardiac ultrasound (echocardiogram) is a critical component of cardiologic imaging. This step is
I wanted to see whether NInfer could run Qwen3.8-27B on an RTX PRO 4000 Blackwell 24GB. Hardware/software: NVIDIA RTX PRO 4000 Blackwell, 24GB 145W power limit CUDA 13.3 NVIDIA driver 610.x Ubuntu LXC
EDIT: Though the issue with q8 kv cache seems to arise from when and how often we run the quantize step, not that kv quantizing can't ever work - see comments for various proposed solutions --- One of
TL;DR: llama.cpp with --load-mode mmap used 21-32 GB of RAM, with ik_llama.cpp using 106-108 GB. -sm tensor killed my prefill, changing to -sm layer went from 36 tps to 135 tps. After that, -ubatch 20
I've got Qwen3.8-Flash-next running on RTX 3090, Ryzen 9 3950X, a PCIe 3.0 motherboard, and 64GB DDR RAM from 2020. IQ4_XS weights, full kvarn5 context, vision on GPU, experts in host RAM, n-grams on
arXiv:2608.27328v1 Announce Type: new Abstract: High similarity between first-visit and return frames does not necessarily show that a video world model remembered the scene; the intervening rollout m
arXiv:2608.27394v1 Announce Type: new Abstract: Retrieved scientific literature can serve as inspiration for both human and AI scientists. Inspiration can take different forms: prior work may directly
arXiv:2608.26644v1 Announce Type: new Abstract: Event cameras capture pixel-level intensity changes with microsecond resolution to produce highly sparse asynchronous data streams. For visual perceptio
arXiv:2608.26418v1 Announce Type: cross Abstract: Modern AI workloads and the hardware that runs them evolve on different timescales: architectural definition precedes volume silicon by years, while t
arXiv:2608.26167v1 Announce Type: new Abstract: Hallucination and abstention benchmarks rarely establish that a model could not have known the correct answer, making it difficult to distinguish approp
We're releasing Qwen3.8-27B quantized with our newest methods, GSQ + RCO. Higher-quality models, same file size, now with the search and the quantizer both learned. What's inside: GSQ (Gumbel-Softmax
arXiv:2608.26833v1 Announce Type: new Abstract: Modern AI has greatly expanded the capabilities of image processing. However, the ready availability of powerful models, public datasets, and benchmark
arXiv:2608.27154v1 Announce Type: new Abstract: Vision Language Models (VLMs) have shown great success in general visual tasks, yet they still struggle to deeply understand text within images. In this
arXiv:2608.26136v1 Announce Type: new Abstract: Sparse autoencoders (SAEs) decompose language-model activations into sparse, interpretable features, and an appealing way to aim them at reasoning is to
arXiv:2608.27033v1 Announce Type: new Abstract: We introduce Riemann-1.0, a fully causal autoregressive World Action Model for embodied intelligence. Riemann-1.0 jointly models multi-view visual obser
Their last version 7.14 was released just a month ago. llama.cpp PR(waiting for approval) for Version 10.0 https://github.com/ggml-org/llama.cpp/pull/27803 Hope this version comes with more boost & im
arXiv:2608.26832v1 Announce Type: new Abstract: Large language models (LLMs) are increasingly applied to specialized domains, where effective use of domain expertise often requires reasoning over comp
arXiv:2608.27135v1 Announce Type: new Abstract: Multimodal foundation models are increasingly used in speech-first assistants that must interpret spoken queries and produce visually grounded decisions
arXiv:2608.27108v1 Announce Type: cross Abstract: Federated Learning (FL) enables privacy-aware distributed training, yet gradient updates remain exploitable: Man-in-the-Middle (MitM) interception exp
arXiv:2607.19326v2 Announce Type: replace Abstract: Low-rank adaptation introduces a static learned update applied identically to every input. The update provides task-level adaptation but does not ex
arXiv:2608.26132v1 Announce Type: cross Abstract: Labeled property graphs combine relational structure with heterogeneous textual and categorical properties attached to both nodes and relationships. C
arXiv:2608.27073v1 Announce Type: new Abstract: Explorable image-to-scene generation is essential for applications in gaming, robotics, and virtual reality. Existing methods based on video diffusion m
arXiv:2608.26432v1 Announce Type: cross Abstract: Voice agents must call tools and hold multi-turn dialogue entirely through speech, yet the dominant paradigm trains them in text. Existing frameworks
Learn how Salesforce used Amazon SageMaker AI Inference Component placement (the SchedulingConfig parameter) to distribute model copies across multiple Availability Zones, meeting their Multi-AZ high
arXiv:2608.26973v1 Announce Type: new Abstract: Pre-training under limited data requires a different view of scaling than web-scale language modeling. With a fixed data budget but relatively abundant
arXiv:2608.26336v1 Announce Type: cross Abstract: Recent advancements in generative models are pushing video generation toward unbounded streaming audio-video generation for real-time interactive worl
arXiv:2608.26437v1 Announce Type: new Abstract: Graph neural networks (GNNs) can exhibit unfair behavior even when sensitive attributes are excluded from node features, because graph topology and mess
arXiv:2608.27367v1 Announce Type: cross Abstract: Joint-Embedding Predictive Architectures (JEPAs) for world modeling typically employ fixed-size Vision Transformer encoders that are over-provisioned
arXiv:2608.26139v1 Announce Type: cross Abstract: Large Language Models demonstrate remarkable syntactic fluency, yet the optimization dynamics governing their acquisition of deep semantic dependencie
arXiv:2603.03230v2 Announce Type: replace-cross Abstract: The electric vehicle routing problem with time windows (EVRPTW) extends the classical VRPTW by introducing battery capacity constraints and ch
arXiv:2608.26923v1 Announce Type: new Abstract: We present TabuLM, the first language model pre-trained on Kinyarwanda tabular data. Kinyarwanda is a morphologically rich Bantu language spoken by over
arXiv:2608.27282v1 Announce Type: cross Abstract: Most single-stage 3D object detectors complete different tasks with the same extracted features. Nevertheless, it is impossible to project features in
arXiv:2608.26971v1 Announce Type: new Abstract: In recent years, image-to-video (I2V) generation models have made remarkable progress in subject consistency and temporal coherence, enabling high quali
arXiv:2505.11326v2 Announce Type: replace-cross Abstract: Vision-language models (VLMs) have shown remarkable progress in offline tasks such as image captioning and video question answering. However,
Post rédigé avec l'aide de GLM5.3 flash Contexte : formateur professionnel d'adultes en France, j'utilise des agents LLM au quotidien et le français qu'ils produisent est mon critère n°1, avant les be
arXiv:2606.07867v2 Announce Type: replace Abstract: Are tool-calling LLM agents equally safe throughout a conversation? We discover they are not: agents are most vulnerable at the very start of a sess
The DeepSeek harness is engineering at its finest. 202K stars on GitHub. Damn! Clearly, they built this harness with future AI needs and capabilities in mind. Everything is a plugin and customizable,
The interns said they'd work late this week. Our MCP and I are just helping The smart home turned into a hobby. Apps, hubs, connectors that break with every firmware update. You gave the robot vacuum
The most popular use case for Claude Tag by far -- on-call. Learn about how we drive Anthropic's on-call with Tag and set it up yourself so you don't get woken up by an alert at 3am that Claude could
arXiv:2608.26235v1 Announce Type: new Abstract: Accuracy-only benchmarking of reasoning-capable large language models misses a central deployment question: when do extended thinking tokens earn their
arXiv:2608.26602v1 Announce Type: cross Abstract: Large software repositories are often beyond model context limits. Training repository knowledge into models is costly and quickly stale, while local
Things getting wild in automated systems engineering. This new work claims to build an AI system that designed and deployed a frontier accelerator. Two human architects wrote the hardware specificatio
arXiv:2601.13029v4 Announce Type: replace Abstract: While Vision-Language Models (VLMs) excel at 2D visual understanding, they remain constrained by 2D-centric paradigm that severely limits genuine 3D
arXiv:2608.26647v1 Announce Type: new Abstract: Full-image objectives in brain magnetic resonance imaging (MRI) super-resolution (SR) can underweight tissue-transition regions affected by the partial-
Hey all, and hello fellow DGX Spark-ers! Today I managed some pretty crazy numbers: 181 tok/s aggregate on 2× DGX Spark on Qwen3.8-Flash-Next at 512Kcontext (2.8M kvc) I hit 181 tok/s aggregate today
Hey everyone, I am the co-founder of Tontaube. My brother and I just released TontaubeV1, a 2.9B-parameter open-weight TTS model focused on expressive speech, long-form generation, and low-latency loc
arXiv:2512.06694v2 Announce Type: replace Abstract: Topic discovery is a fundamental technique for text mining that identifies abstract topics within large document collections. A recent approach to t
arXiv:2608.26417v1 Announce Type: cross Abstract: Metasurface design increasingly requires fast models that can operate across structurally distinct device families, rather than retraining a separate
After getting tired refreshing and searching sub reddits for qwen 3.8 35 a3b i decided to give qwen 3.8 flash next a try . So I dual booted and built llama cpp I am able to hit 6-7 tps constantly insi
arXiv:2608.27182v1 Announce Type: new Abstract: LLM agents are increasingly applied to anomaly detection and root-cause analysis in time-series observations collected from real-world systems; however,
arXiv:2603.25489v2 Announce Type: replace Abstract: Recent strategies for low-resource machine translation rely on LLMs to generate synthetic data based on text in higher-resource languages. We revisi
arXiv:2608.27128v1 Announce Type: new Abstract: Long-context inference is bottlenecked by the memory footprint of the key-value (KV) cache, especially for small models under tight resource budgets. Ex
arXiv:2608.27351v1 Announce Type: new Abstract: Evolution Strategies (ES) have recently emerged as a memory-efficient post-training paradigm for LLM reasoning. However, the optimization behavior of ES
arXiv:2608.27070v1 Announce Type: cross Abstract: To mitigate catastrophic forgetting in downstream continual learning (CL) for large language models (LLMs), existing methods typically constrain param
What's Changed Ollama's app now follows the system appearance again, restoring dark mode support Fixed the macOS app to properly hand off to an already-running instance instead of starting a second on
arXiv:2608.26382v1 Announce Type: new Abstract: Pathology vision-language models are advancing rapidly, yet existing benchmarks remain focused on human tissue, particularly oncology, leaving non-human
arXiv:2608.27199v1 Announce Type: new Abstract: Vision generative artificial intelligence (AI) has emerged as one of the most rapidly advancing areas of deep learning. The explosion of multimodal mode
We are rolling out GLM-5.3 on Ollama. Private. Fast. US and Europe hosted. No data retention. Try Ollama's GLM-5.3-Flash as we bring it GLM-5.3 online. Super fast. Claude Code: ollama launch claude --