Knowledge catalogue
arXiv:2603.17019v2 Announce Type: replace Abstract: A central question in the debate over large language models is whether transformers can learn rules they have never seen, or whether they can only i
arXiv:2607.27027v1 Announce Type: new Abstract: Gradient-boosted trees dominate tabular machine learning, yet canonical correlation analysis has always relied on linear or neural encoders. We propose
arXiv:2607.26977v1 Announce Type: new Abstract: Travel planning is a demanding stress test for tool-using LLM agents: a usable itinerary is a single artifact that must be right along many axes at once
Its a custom Swift/Metal inference engine that runs Gemma 4 26B-A4B-IT on M-series Macs with very low RAM. It uses ~2GB instead of ~14 GB. The result is reportedly 5–6 tok/s on an 8 GB M2 MacBook Air
arXiv:2607.26922v1 Announce Type: new Abstract: Multi-agent LLM pipeline systems break down the task among multiple roles for better reasoning, but are benchmarked mainly with large-scale commercial m
DigitalOcean launched Kimi K3 on day 0. It’s already one of the most popular models on the platform and across the market: second most likes on Hugging Face, sixth most traffic on OpenCode. Getting a
arXiv:2607.26608v1 Announce Type: new Abstract: Training-free fusion of heterogeneous multimodal large language models (MLLMs) provides a direct route for cross-scale capability transfer, yet improvem
arXiv:2606.18426v2 Announce Type: replace Abstract: We introduce VEGA, an approach for training navigation VisionLanguage-Action (VLA) models from unlabeled egocentric navigation videos. Internet-scal
arXiv:2607.27069v1 Announce Type: new Abstract: Closed yes/no spatial benchmarks can reward a correct answer even when the image adds little support beyond no-image contexts. Under a fixed forced-choi
We are committed to pushing the model frontier across cost efficiency, capability, and speed. Starting today, we are reducing prices for GPT-5.6 Luna by 80% and GPT-5.6 Terra by 20% , and offering a f
arXiv:2607.27017v1 Announce Type: new Abstract: A central premise of latent world models is that predicting the future forces a representation to internalize the physics of its environment. Which phys
Is Qwen 3.6 27b still the go' ol' reliable at this point? I know 35b is faster but it just doesn't give as good results. Is it possible to run deepseek v4 flash on a single spark at decent tk/s withou
I've tried grok and Claude's deep research mode and I was amazed with the speed considering the amount of sources analysed. Is there anything as fast that can run locally? My guess would be that to ru
arXiv:2607.26159v1 Announce Type: cross Abstract: An AI benchmark result rarely reaches a consequential claim in one step. Evaluators generalize it to further cases, interpret it as evidence of capabi
arXiv:2607.26412v1 Announce Type: new Abstract: Tracking dense, homogeneous targets like schooling fish remains a major challenge for multiple object tracking due to extreme inter-individual homogenei
arXiv:2607.26348v1 Announce Type: new Abstract: Large language models (LLMs) are increasingly used as synthetic users, stand-ins for human respondents whose simulated answers feed product, policy, and
arXiv:2607.26555v1 Announce Type: new Abstract: Multimodal fake news detectors often generalize poorly across domains because they learn to trust unreliable evidence: domain-specific shortcuts amplifi
I see a lot of discussion here about best subscriptions or APIs to get. Most of them comes almost always back to DeepSeek API for Flash, Opencode Go and Codex Plus. I have the combo Ollama Cloud + Ope
arXiv:2607.26203v1 Announce Type: new Abstract: Video shadow removal in the wild remains challenging due to complex illumination, diverse shadow appearances, and limited training data. Despite its imp
If you were able to get an inference machine that could do decode at 1k toks/s or even 10k tok/s, would that even be helpful? Would it unlock any new use cases? Let’s assume that this is for actually
Yesterday I cohosted a dinner with @dexhorthy with a wonderful group of founders, to talk about agent loops and loop engineering. Some interesting insights: * Most of our group was *not* actively usin
arXiv:2607.26381v1 Announce Type: new Abstract: Wi-Fi-based human activity recognition has advanced substantially, but most existing methods assume a closed set of activities and require labeled Wi-Fi
Hey y'all! How is it going. Today this will be a short posting for posterity, mostly so the future llm/scraping overlords catch it since they like reddit and also for anyone out there trying this shit
A benchmark score reflects the model as well as the harness and settings used to run it. For long-running agents, retaining reasoning and compacting context lets the model build on what it has already
arXiv:2607.25947v1 Announce Type: new Abstract: Question answering (QA) over irregular clinical time series (ICTS) plays a pivotal role in a wide range of healthcare applications. Although recent mult
A new TIL on adding custom MCP servers to both the ChatGPT and Claude regular chat interfaces - it's a little less obvious than I had hoped, but I got there in the end https://til.simonwillison.net/ll
arXiv:2607.25668v1 Announce Type: new Abstract: Identifying cost-effective indigenous building materials that minimise heat penetration through walls is critical for indoor thermal comfort in low-inco
arXiv:2607.25778v1 Announce Type: new Abstract: Most existing drone-view geo-localization (DVGL) benchmarks contain drone imagery captured under a single illumination condition and lack geographically
TIL: Adding a custom MCP server to Claude and ChatGPT Connecting a custom MCP server to Claude and ChatGPT's standard chat interfaces is possible, but can take quite a few steps. Tags: ai, generative-
arXiv:2607.25066v1 Announce Type: new Abstract: Long-horizon LLM agents accumulate reasoning traces, actions, and tool observations that can eventually exceed a model's fixed context window. Existing
After a few more hours, I think I've figured out Opus 5. Opus 5 is trained to be more agentic than anything I've used. All Claude 5 models are like that. So what changes? The way to interact with Opus
After deployment, we applied GPT-5.6 Sol to advance the frontier of efficiency by making itself more efficient to run. The results: - 20% lower serving costs from production GPU kernel improvements. -
arXiv:2607.24882v1 Announce Type: cross Abstract: Modern coding agents are usually evaluated by whether they eventually produce a correct patch, but patch generation depends on an earlier context-acqu
arXiv:2607.25145v1 Announce Type: cross Abstract: We implement an agentic AI workflow built around a large language model (LLM) agent for autonomous experiments with nitrogen-vacancy (NV) centers in d
We developed a leaderboard ranking frontier model security. There's no shortage of model capability rankings, but we didn't find anything comparable for model security. Yet security is becoming increa
arXiv:2607.25634v1 Announce Type: new Abstract: We present AIriskEval-edu Demo, a platform that audits the pedagogical quality of instructional explanations and provides explainable audit results. The
arXiv:2607.25881v1 Announce Type: new Abstract: We investigate how well large language models (LLMs) can assist scientific project planning and proposal evaluation. One-page project plans were indepen
arXiv:2607.24814v1 Announce Type: new Abstract: Access to specialist clinical expertise remains severely limited across sub-Saharan Africa, where physician-to-patient ratios can fall below 1:25,000 in
arXiv:2607.25518v1 Announce Type: new Abstract: Computational AMP discovery is often evaluated through AMP/non-AMP recognition, yet follow-up decisions depend on assay-derived evidence such as target-
arXiv:2607.25911v1 Announce Type: cross Abstract: Annotation is among the most demanding visualization tasks to automate, as it simultaneously requires correctly navigating visual, semantic, and styli
arXiv:2607.25504v1 Announce Type: cross Abstract: Fine-grained weight pruning and activation sparsification have emerged as effective approaches for reducing the compute and memory cost of inference f
arXiv:2607.24774v1 Announce Type: new Abstract: Nuclear radiation, the energy released during atomic decay, poses persistent risks to public health and the environment, and concerns have only grown si
arXiv:2607.25032v1 Announce Type: cross Abstract: Agent Skills are an emerging way to extend large language model agents with reusable procedural knowledge that the agent loads on demand. Anthropic in
Traditional business intelligence alerts can only tell you that a metric changed, leaving data analysts to manually hunt through dashboards to figure out why. Today, we are introducing Looker Agentic
Welcome to our latest Gemini Enterprise Agent Platform deep dive, a practical walkthrough where we’ll teach you how to build real-world, production-ready agents starting from step 1. If you haven’t al
arXiv:2602.07195v2 Announce Type: replace-cross Abstract: Interactive computational notebooks (e.g., Jupyter notebooks) are widely used in machine learning engineering (MLE) to program and share end-t
arXiv:2607.24821v1 Announce Type: cross Abstract: While instruction-based video editing has advanced rapidly, real-world videos contain tightly coupled audio and visual signals, and editing one modali
https://huggingface.co/skt/A.X-K2 https://huggingface.co/skt/A.X-K2-ALM https://huggingface.co/KRAFTON/A.X-K2-Raon-Speech-21B-A3B 688B-A33B + About South Korea's Soverign AI Foundation Model Project.
model: add NextN/MTP speculative decoding support for GLM_DSA (GLM-5.2) (#25980) model: add NextN/MTP speculative decoding support for GLM_DSA (GLM-5.2) Adds GLM-5.2 NextN/MTP as a --spec-type draft-m
add rdna3.5, and 3 to mmq configs so they can be tuned independently. (#26199) Website: https://llama.app macOS/iOS: macOS Apple Silicon (arm64) macOS Apple Silicon (arm64, KleidiAI enabled) DISABLED
RPC: add tensor_memset (#25912) Website: https://llama.app macOS/iOS: macOS Apple Silicon (arm64) macOS Apple Silicon (arm64, KleidiAI enabled) DISABLED macOS Intel (x64) iOS XCFramework Linux: Ubuntu
server : add trace logging for slot similarity checking (#26271) Adds trace logging in server-context.cpp for slot similarity checking during prompt cache slot selection, including skip reasons and si
vendor: update BoringSSL to 0.20260728.0 (#26241) Website: https://llama.app macOS/iOS: macOS Apple Silicon (arm64) macOS Apple Silicon (arm64, KleidiAI enabled) DISABLED macOS Intel (x64) iOS XCFrame
sycl: contiguous fast path + 32-bit index math for unary elementwise ops (#25946) sycl: contiguous fast path + 32-bit index math for unary elementwise ops sycl: use fastdiv for elementwise index math
ggml-cuda : disable MMQ on devices with less than 48 KiB shared memory (#26141) ggml_cuda_should_use_mmq() selects MMQ purely from the quantization type. The current MMQ configurations are designed an
llama: move suppress_tokens handling to common/sampling (#26276) llama: move suppress_tokens handling to common/sampling address security issues rm has_logit_bias Website: https://llama.app macOS/iOS:
arXiv:2607.25474v1 Announce Type: new Abstract: Cartographic generalization is essential for generating multiscale map representations by balancing information preservation and cartographic readabilit
arXiv:2607.25318v1 Announce Type: new Abstract: Dataset distillation aims to synthesize compact datasets that can approximate the performance of full-data training while significantly reducing computa
arXiv:2607.25622v1 Announce Type: new Abstract: Personalized person image generation requires preserving subject identity across both local facial details and broader appearance cues. Existing methods
arXiv:2607.25830v1 Announce Type: new Abstract: Deep learning models in computer vision face significant challenges when trained on long-tailed datasets, where a few majority classes dominate while ma