Knowledge catalogue
Hiya, NInfer is amazng for Qwen, but lacking for real-world-use. As adoption of issues/pr's was not really what I needed, I created a fork and hit it for this week with 3 concurrent claude code sessio
This is very similar to how we think about context management in Deep Agents: compaction should reduce what is sent to the model, not erase the underlying work history. In Deep Agents, the logic is: -
arXiv:2609.01772v1 Announce Type: new Abstract: Meme understanding goes beyond recognizing visual content or literal text; it requires implicit cultural knowledge and pragmatic inference that most vis
This was my first serious attempt at tuning a local LLM. I started because Qwen3.8-27B IQ3 was fast on my RTX 5080 but the coding quality disappointed me, and the Q4 profiles I tried in LM Studio were
arXiv:2602.18493v2 Announce Type: replace-cross Abstract: Long-context LLMs and Retrieval-Augmented Generation defer state tracking and evidence consolidation to query time, which is brittle when fact
arXiv:2608.30689v1 Announce Type: new Abstract: Autoregressive large models have recently advanced Text-to-SVG generation from simple icons to complex, long-context graphics, yet standard autoregressi
arXiv:2608.30315v1 Announce Type: cross Abstract: Token embeddings are the basic representational units that connect discrete tokens with continuous computation in language models. Although modern lan
arXiv:2608.29847v1 Announce Type: cross Abstract: Text-to-image models learn associations between concepts - in the case of this paper, people's professions, which we refer to as roles - and visual at
arXiv:2510.03204v2 Announce Type: replace Abstract: Web agents powered by large language models (LLMs) must process lengthy web page observations to complete user goals; these pages often exceed tens
arXiv:2512.16391v2 Announce Type: replace-cross Abstract: Attention is the dominant source of latency during long-context LLM inference, an increasingly popular workload with reasoning models and RAG.
arXiv:2607.17486v2 Announce Type: replace-cross Abstract: As large language models (LLMs) process increasingly longer prompts, computation and KV-cache memory costs have emerged as major bottlenecks i
arXiv:2608.28070v1 Announce Type: new Abstract: Automated detection of surface micro-defects on industrial components, such as copper tubes, is critically important for quality assurance but remains c
I used the quantized n-gram to INT4, it's 32 GB, memory-mapped from disk. I confirmed that it works great on 150-160k context, and i was watching all the time my VRAM usage while doing single thread l
Setup: MacBook Pro M5 Max, 128 GB unified, macOS 26.5.2 · llama.cpp b10686 (Metal, 12 threads, batch 2048, flash-attn, kv-unified, ngram-mod spec decode) · Qwen3.8-Flash-Next UD-Q2_K_XL (Unsloth), 78.
Hey! I forked NInfer (a from-scratch C++20/CUDA inference engine for Qwen models) and added two things: tensor-parallel across two GPUs, and YaRN ×4 rope scaling. Together they let Qwen3.8-27B NVFP4 r
arXiv:2608.27036v1 Announce Type: new Abstract: Large Language Models (LLMs) can be trained to perform chain-of-thoughts reasoning in order to improve the reliability of their responses. In this work,
arXiv:2608.25657v1 Announce Type: cross Abstract: Large language models (LLMs) excel at programming, but not when the task fixes the target language: prompted with a grammar rare in their training dat
arXiv:2608.25655v1 Announce Type: new Abstract: Conversations with chat assistants increasingly span many topics in a single long-running thread, challenging memory systems. Existing long-context and
arXiv:2608.24956v1 Announce Type: new Abstract: Data-driven methods are widely used in landslide susceptibility mapping (LSM) because they can effectively model the complex relationships between lands
arXiv:2608.25096v1 Announce Type: new Abstract: The growing adoption of large language models (LLMs) has raised increasing concerns about the energy consumption and environmental impact of inference.
arXiv:2608.23855v1 Announce Type: new Abstract: We propose ICI-Time, a novel framework that reframes time series forecasting as a visual inpainting task, leveraging the generalisation power of large v
arXiv:2608.23330v1 Announce Type: new Abstract: Video understanding requires intelligent agents to transcend mere recognition of visual facts and comprehend the underlying intents behind human actions
arXiv:2608.21242v1 Announce Type: new Abstract: As conversational companions, large language models (LLMs) often have access to users' emotional states. We study how this affective context modulates L
arXiv:2608.21229v1 Announce Type: new Abstract: Omnimodal generation is central to a wide range of content creation and editing applications. In-context conditioning is essential to this paradigm. It
arXiv:2608.20427v1 Announce Type: cross Abstract: Dense causal attention remains expensive at long context even when implemented with highly optimized exact kernels. We study BF1, a deterministic bloc
Spent a while treating layer placement, KV format and llama.cpp itself as experimental variables. 159 logged experiments. Numbers first, caveats after. Hardware: AMD Ryzen AI MAX+ 395 (Strix Halo, 128
arXiv:2608.17950v1 Announce Type: new Abstract: Large Language Models (LLMs) demonstrate remarkable multi-hop reasoning capabilities over long contexts, yet the internal mechanisms enabling these dist
arXiv:2608.17829v1 Announce Type: cross Abstract: LLMs increasingly rely on external contexts, such as pre-defined system prompts or retrieved documents, to improve generation quality. However, proces
arXiv:2608.15580v1 Announce Type: new Abstract: Reliable endoscopic polyp reporting requires integrating quantitative lesion sizing, standardized Paris classification, and clinically meaningful morpho
arXiv:2608.15879v1 Announce Type: new Abstract: Large language models (LLMs) have shown strong performance in text generation tasks, yet their effectiveness on headline generation remains sensitive to
arXiv:2608.12615v1 Announce Type: cross Abstract: In-vehicle music can serve as an adaptive interface to enhance driver experience, attention, and well-being. We present Drive-to-Music, a context-awar
arXiv:2606.20077v2 Announce Type: replace-cross Abstract: Visual tokens enter Large Language Models (LLMs) as raw, foreign signals. How they are transformed into meaningful representations and interac
arXiv:2608.13283v1 Announce Type: new Abstract: Understanding motion in daily living requires context beyond kinematics, because similar inertial patterns during activities of daily living (ADLs) can
arXiv:2608.10343v1 Announce Type: new Abstract: While deep learning-based denoising has become widely adopted in low-dose CT, conventional models use generic architectures designed for natural images,
arXiv:2608.08212v1 Announce Type: new Abstract: In-context learning (ICL) can induce emergent misalignment (EM), where narrow misaligned examples alter answers to unrelated questions. Existing prompts
Heeeey all! I just completed some fun tests with Muse Glimmer, I thought I'd let you know. In fact, the summary below was written by Muse itself! I ran a 2× DGX Spark cluster and got Meta's day-old Mu
Knowledge is rather abstract and philosophical. So what does it mean from an agentic context? How does providing the right knowledge at the right time to an agent increase its accuracy, task completio
Available context length with and without the patch: Model: QWEN 27B ROCm stock patched Vulkan stock patched IQ4_XS Pure, single 16GB GPU 19.456 76.032 68,352 78,592 Q6_K_L on 16GB + 12GB 64,256 149,2
arXiv:2608.06105v1 Announce Type: cross Abstract: Artificial Intelligence (AI)-assisted navigation can help Arctic shipping adapt to rapidly changing sea-ice conditions, but reliable deployment requir
LabyrinthBench measures the thing that actually kills long agent runs — whether a model can still use what it learned twenty turns ago — deterministically, with no LLM judge, on your own hardware, wit
arXiv:2608.06377v1 Announce Type: cross Abstract: Language models increasingly condition their answers on external signals, and a single misleading one can turn a correct answer wrong. The obvious rem
arXiv:2608.05707v1 Announce Type: new Abstract: Frame selection is essential for applying Large Multimodal Models (LMMs) to long videos due to severe frame redundancy and limited context windows. Sinc
arXiv:2608.03015v1 Announce Type: cross Abstract: Pretrained large language models (LLMs) have demonstrated in-context learning (ICL) capabilities for numerical inference over sequences serialized as
arXiv:2511.10696v3 Announce Type: replace-cross Abstract: Sparse attention is crucial in long-context Transformers, which restricts each token to a limited neighborhood and thereby reduces the quadrat
arXiv:2607.29595v1 Announce Type: new Abstract: Ultra-high-definition (UHD) image restoration must balance the aggregation of spatially recurring degradation cues with the preservation of localized im
arXiv:2604.08342v2 Announce Type: replace Abstract: Long context egocentric video understanding has recently attracted significant research attention, with augmented reality (AR) highlighted as one of
arXiv:2607.25291v1 Announce Type: new Abstract: The quadratic cost of self-attention makes long-context inference prohibitively expensive, and proxy-based block-sparse attention has become a practical
arXiv:2607.22566v1 Announce Type: new Abstract: MedLoCoMo is a Medical Long-Context Memory benchmark for patient-specific clinical reasoning over multi-admission medical dialogue. Existing medical QA
Great technical paper from Google. Great read on why context beats scale for agents working against unfamiliar APIs. (bookmark it) GPU kernel optimization has KernelBench to hillclimb on. TPUs had not
Most graph-based LLM interfaces use a canvas as a visual layer over what is still a linear chat. I wanted the graph itself to determine what Ollama receives. ThoughtDAG has one rule: wires are the con
https://preview.redd.it/yw41s1jikefh1.png?width=1942&format=png&auto=webp&s=3a180ae6443c1db9f7b0ce621533a4b2aa553921 Hi, I've been running Ollama on my Unraid server since the llama2 era. I use to be
arXiv:2607.21427v1 Announce Type: new Abstract: Discrete flow matching provides a flexible framework for generative modeling on discrete structures. However, the standard factorized training objective
arXiv:2607.19595v1 Announce Type: cross Abstract: Large language model (LLM) agents are vulnerable to security risks, such as prompt injection attacks from untrusted context that manipulate downstream
We built this experience based on feedback from early testers and physicians. With your permission, ChatGPT can use relevant context you’ve connected in Health across your conversations — to help you
arXiv:2607.13712v1 Announce Type: cross Abstract: Despite the rapid progress of Multimodal Large Language Models (MLLMs), they still suffer from untruthfulness issues, such as visual hallucinations, c
arXiv:2601.02023v2 Announce Type: replace-cross Abstract: As Large Language Models (LLMs) increasingly utilize massive context windows as working memory for autonomous tasks, their reliability fluctua
arXiv:2510.18939v2 Announce Type: replace Abstract: Long-horizon agentic search requires iteratively exploring the web over long trajectories and synthesizing information across many sources, enabling
arXiv:2507.05240v2 Announce Type: replace-cross Abstract: Vision-and-Language Navigation (VLN) in real-world settings requires agents to process continuous visual streams and generate actions with low
arXiv:2607.07409v1 Announce Type: new Abstract: Speculative decoding accelerates LLM inference by drafting multiple tokens and verifying them in parallel. Block-parallel drafters such as DFlash furthe
arXiv:2607.07144v1 Announce Type: new Abstract: The key-value (KV) cache dominates the memory cost of long-context autoregressive inference, and a growing body of work compresses it through quantizati