Qwen 3.8 27b is strong even at Q3_xxs
So usually I avoid Q3 quants because I have had bad experiences with it, models were usually too degraded, so the smallest I normally do is Q4, since I only have rtx 4060 ti 16gb. But since there hasn
Knowledge catalogue
So usually I avoid Q3 quants because I have had bad experiences with it, models were usually too degraded, so the smallest I normally do is Q4, since I only have rtx 4060 ti 16gb. But since there hasn
arXiv:2608.19658v1 Announce Type: new Abstract: This study demonstrates a rationally enriched Chebyshev (REC) trunk for deep operator network (DeepONet) surrogate models of singularly perturbed and hi
arXiv:2608.19689v1 Announce Type: new Abstract: LLM-based social simulation is a promising complement to traditional methods such as surveys and behavioral experiments. A core question is how to evalu
arXiv:2608.19885v1 Announce Type: cross Abstract: Streaming systems that maintain a pool of expert models must repeatedly decide whether to reuse an existing expert for arriving data, spawn a new one,
arXiv:2608.19238v1 Announce Type: cross Abstract: Spiking Transformers model token interactions primarily through spiking self-attention (SSA). However, binary query and key representations map contin
arXiv:2412.06265v3 Announce Type: replace Abstract: Deep tabular models should ideally balance predictive performance, parameter efficiency, and robustness to imperfect learning signals---properties t
This is very nice work from NVIDIA. Like all high-performing approaches on ARC-AGI-3, it uses deep learning-guided on-the-fly synthesis of symbolic world models, i.e. navigating the world by generatin
arXiv:2608.19825v1 Announce Type: cross Abstract: Medical image captioning is a technique that accelerates early-stage diagnostic workflows and enhances the interpretability of medical diagnostic AI s
arXiv:2512.18046v2 Announce Type: replace Abstract: Unmanned Aerial Vehicles, commonly known as, drones pose increasing risks in civilian and defense settings, demanding accurate and real-time drone d
tldr: went from 16~ t/s to 27~ t/s generation. got my usable context up from 220k to the full 262k without sacrificing anything. prefill also increased from 376 to 573 command I ended up with, fwiw: l
arXiv:2608.18731v1 Announce Type: cross Abstract: Medical image segmentation is essential for clinical workflows such as treatment planning and disease assessment. While specialist tools like TotalSeg
arXiv:2608.18183v1 Announce Type: new Abstract: Visual on-policy distillation (OPD) improves the training of compact visual autoregressive models by learning from trajectories generated by the current
arXiv:2608.18888v1 Announce Type: new Abstract: The rapid advancement of Natural Language Generation (NLG) has made the reliable evaluation of generated text increasingly critical, as these systems, s
arXiv:2608.18101v1 Announce Type: new Abstract: Longitudinal text streams exhibit topic birth and death, but also discrete structural reorganizations in which themes split into subtopics or merge into
arXiv:2511.22842v3 Announce Type: replace-cross Abstract: Causal machine learning (Causal ML) aims to answer 'what if' questions using machine learning algorithms, making it a promising tool for high-
arXiv:2608.18307v1 Announce Type: new Abstract: Current evaluation of computer-use agents is split between long-horizon workflow benchmarks and atomic GUI-grounding tests. This leaves an under-instrum
arXiv:2608.18613v1 Announce Type: new Abstract: Cyber threat intelligence (CTI) is increasingly consumed not by human analysts but by LLM agents that compose multi-step investigations at query time. T
arXiv:2605.06185v2 Announce Type: replace Abstract: Large vision-language models perform well on short- and medium-length video understanding but still struggle to maintain coherent event memory and r
arXiv:2608.18900v1 Announce Type: new Abstract: As AI systems are increasingly deployed in safety-critical application domains (e.g., autonomous driving), associated risks increase too. Deep learning
Hey LocalLlama, Henry from Cactus here! When we trained Needle 2, I had a strict rule to not expose the model to any data sample that remotely felt like these benchmarks. It seemed over-the-top, but b
arXiv:2608.18733v1 Announce Type: cross Abstract: We present Flama, an open-source Python framework for developing and deploying production-ready web APIs, machine learning services, and large-languag
arXiv:2608.18097v1 Announce Type: new Abstract: We present FrenchNews-7, a cross-publisher France-based French-language news editorial desk classification benchmark combining a large multi-outlet corp
arXiv:2608.18155v1 Announce Type: cross Abstract: Quantum machine learning (QML) for network intrusion detection (NIDS) is routinely reported to reach near-perfect accuracy, yet the most rigorous stud
I've been building AgentGraph Studio, a visual editor for designing multi-agent workflows and exporting them as runnable Python code. One thing I wanted to verify was whether the generated code actual
arXiv:2608.18412v1 Announce Type: new Abstract: Effective sign language (SL) acquisition is crucial for deaf children, yet 95% are born to hearing parents who often lack proficiency in SL. SL recognit
AntLing has released the full six-checkpoint matrix for the Ling-3.0 base model. tiny: pretrained, mid-trained, WSM-merged flash: pretrained, mid-trained, WSM-merged The concrete artifact is six separ
arXiv:2608.18571v1 Announce Type: new Abstract: Sleep stage classification from single-channel electroencephalography (EEG) is essential for wearable and home-based sleep monitoring. However, many dee
I currently have these models available through Ollama/cloud: And I’m using Claude, Codex, OpenCode, and Ollama as my coding/agent tools. I want to learn professional vibecoding — not just asking AI t
arXiv:2608.18090v1 Announce Type: cross Abstract: Inside a modern language model sits a single internal direction that tracks how positive or negative a sentence feels. We show how to find this valenc
arXiv:2608.18438v1 Announce Type: cross Abstract: Modern mental healthcare faces a critical shortage of senior supervisory oversight, leading to a 'supervision gap' where novice therapists manage high
I benchmarked Qwen3.8-27B on MathArena/aime_2026 dataset, comparing BF16 and FP8 weights at medium and xhigh reasoning effort. Interesting findings are: quantized FP8 xhigh is better than BF 16 medium
arXiv:2608.18271v1 Announce Type: new Abstract: On-policy self-distillation (OPSD) trains a student on its own responses using token-level supervision from the same model conditioned on privileged ref
arXiv:2608.18190v1 Announce Type: new Abstract: Domain adaptation is widely used to make neural networks trained on simulations applicable to experimental data. Its premise is that the two domains dif
arXiv:2608.18379v1 Announce Type: cross Abstract: When every candidate is wrong, correct-candidate selection is unavailable, yet the aggregation call can still solve the problem afresh. A correct aggr
arXiv:2608.18102v1 Announce Type: new Abstract: The widespread adoption of large language models (LLMs) has intensified the demand for principled methods to distinguish human from machine-generated te
Web agent economics are set by inference volume and each step is one inference call. A simple task (extract a field, fill a form, navigate a site) is 10 to 20 calls. A multi-site workflow runs into th
Hello, any params I missing or mis-tuning for my coding model ? FROM ./models/unsloth/Qwen3.8-27B-Q5_K_M.gguf # Generation parameters optimized for logical reasoning and coding tasks PARAMETER tempera
Aloha! 🌺Introducing Ornith-1.5, a family of open-source LLMs spanning 9B Dense, 35B MoE, and 397B MoE, trained with self-improving strategies. It achieves state-of-the-art performance among open-sourc
I have been using local models on/off for like 2 years or so but never really used them extensively because the closed ones were always much better. Once Qwen 3.8 27B was released I decided to give it
arXiv:2608.17638v1 Announce Type: new Abstract: What a reasoning model writes is only a partial record of the process that produces it. We introduce a two-level internal readout for mixture-of-experts
I am so tired of the PR. How Anthropic's new results post would read without the PR: Claude orchestrated open-source protein design models, PXDesign, RFdiffusion, Genie, BoltzGen, from a 30k-token exp
Edit: Title says 134 tps, it's actually 138 -- keep in mind my 3090 is power limited to 250w. Three days ago I released a hyper-optimized Qwen3.8-27B inference engine for an RTX 3090 (82 tps single re
arXiv:2608.17883v1 Announce Type: new Abstract: The availability of high-quality paired data is essential for training learning-based image demoireing models. However, it remains challenging for exist
arXiv:2608.17279v1 Announce Type: new Abstract: This report presents a two-stage, training-free solution for the MeViS-Text track of the 8th LSVOS Challenge. The task requires a model to localize and
arXiv:2608.17561v1 Announce Type: new Abstract: The health of marine ecosystems is a critical indicator of global environmental change, yet the physical constraints of underwater observation and the i
arXiv:2603.24126v2 Announce Type: replace Abstract: When language models are trained by reinforcement learning (RL) to write probabilistic programs, they can artificially inflate their marginal-likeli
arXiv:2608.17386v1 Announce Type: new Abstract: Foundation-model policies for robotic manipulation are advancing rapidly on task success, but rigorous evaluation of whether they succeed safely is stil
arXiv:2608.17687v1 Announce Type: new Abstract: Despite their widespread use, Large Language Models (LLMs) remain limited by a fundamental problem: the generation of plausible but false content, known
arXiv:2608.17246v1 Announce Type: new Abstract: This article investigates several physics-informed and hybrid machine learning strategies that incorporate physics knowledge in experimental data-driven
arXiv:2608.18008v1 Announce Type: cross Abstract: Combining large language models with reinforcement learning is increasingly explored, yet the theoretical status of LLM-derived reward signals is ofte
arXiv:2608.16970v1 Announce Type: cross Abstract: LLM-based code generation is now embedded in mission-critical pipelines, but defenses against vulnerable output remain post-hoc -- static analyzers, f
arXiv:2608.17379v1 Announce Type: cross Abstract: We introduce PTXBench, a benchmark for evaluating and adapting large language models (LLMs) to use architecture-specific PTX for GPU kernel optimizati
arXiv:2608.17288v1 Announce Type: new Abstract: GPT attention measures token compatibility through dot-product similarity. This mechanism is simple, effective, and memory-efficient. But it does not ex
arXiv:2608.17475v1 Announce Type: new Abstract: Vision foundation models have recently advanced multi-modal salient object detection (MSOD) through parameter-efficient tuning and prompt learning. Howe
arXiv:2608.17164v1 Announce Type: new Abstract: Textual context such as news, reports, and logs can provide valuable signals for time series forecasting, especially when future dynamics are driven by
arXiv:2608.17209v1 Announce Type: cross Abstract: End-to-end vision-language-action (VLA) and world-action models offer an elegant route to general-purpose robotics, but their reliability is bounded b
arXiv:2608.17153v1 Announce Type: new Abstract: Retrieval-Augmented Generation (RAG) has significantly enhanced the performance of large language models (LLMs), yet these systems remain vulnerable to
arXiv:2510.06039v2 Announce Type: replace-cross Abstract: Reliable evaluation of knowledge-grounded Large Language Models (LLMs) in Chinese requires resources that explicitly align Chinese-language te
arXiv:2608.14750v1 Announce Type: cross Abstract: Left ventricular ejection fraction (LVEF) estimation (Task 1), global longitu-dinal strain (GLS)-based dysfunction classification (Task 2), and early
arXiv:2608.15939v1 Announce Type: new Abstract: Stateful language agents assume a rejected branch can be taken back by clearing it from the application transcript. We show this breaks when the serving