Knowledge catalogue
Fred Gao / Inside China: In a leaked four-hour investor talk, DeepSeek's Liang Wenfeng says the main US-China gap is compute access, Nvidia's CUDA moat is disintegrating, and more — In a rare four-hou
If you're running Laguna S 2.1 on llama.cpp and hitting thinking loops because it won't close its </think> tags, you might want to look at your quant before you spend too much time tweaking settings.
arXiv:2602.00462v5 Announce Type: replace-cross Abstract: Transforming a large language model (LLM) into a vision-language model (VLM) can be achieved by mapping the visual tokens from a vision encode
arXiv:2607.19889v1 Announce Type: new Abstract: Understanding instrument-tissue interactions is essential for context-aware surgical AI and autonomous robotic surgery. Pretrained vision-language model
arXiv:2607.18924v1 Announce Type: new Abstract: Recent advances in image-to-video generation have improved visual realism, making physically grounded and controllable dynamics an important step toward
Many still debate open vs closed, and compare cost per token (accounting metric). Better to shift attention to cost per successful task. @seldo and the team at @arizeai did so across 2,400 runs. Concl
arXiv:2607.19826v1 Announce Type: new Abstract: Action Quality Assessment (AQA) aims to objectively evaluate performance quality from action videos. Most existing methods follow a ``one-by-one'' parad
arXiv:2607.17967v2 Announce Type: replace Abstract: Monocular geometry estimation has recently achieved impressive performance across diverse scenes. However, state-of-the-art models still face notabl
arXiv:2607.19288v1 Announce Type: new Abstract: Test-time scaling offers a promising method to improve the inference performance of Vision-Language Models (VLMs) without additional training. Existing
arXiv:2607.19624v1 Announce Type: new Abstract: The allocation of visual attention by pathologists during cancer diagnosis is a highly selective process that critically shapes the information extracte
arXiv:2607.18448v1 Announce Type: cross Abstract: Pathology report generation from whole-slide images (WSIs) is a rapidly growing multimodal learning problem, yet progress is difficult to measure beca
arXiv:2607.19653v1 Announce Type: cross Abstract: Large language model (LLM) agents now perform well on correctness-oriented repository-level tasks, including SWE-Bench issue resolution and feature im
arXiv:2607.17820v2 Announce Type: replace Abstract: Training-free few-shot adaptation methods have gained significant attention recently in the context of Vision-language Models (VLMs). Yet, current b
arXiv:2605.05598v2 Announce Type: replace Abstract: The proliferation of large language models (LLMs) in educational settings has paradoxically undermined the cognitive processes they purport to suppo
arXiv:2607.19408v1 Announce Type: new Abstract: Using Evolutionary Strategies (ES) for fine-tuning large language models is attractive because it is memory-efficient, parallel, and compatible with bla
arXiv:2607.11557v2 Announce Type: replace Abstract: Knowledge distillation (KD) typically relies on the fixed perspective of a single teacher, limiting the diversity of supervisory signals. While mult
Some thoughts on the purported distillation timeline. 1. You don't need that many Fable tokens to distill on top of the alleged ~50b Anthropic claimed Moonshot AI distilled in Feb. ~10b, about $60k wo
arXiv:2607.19518v1 Announce Type: new Abstract: Sophisticated Inference is a variant of active inference often associated with recursive belief modeling and tree search. We argue that its central comp
Prime Intellect Lab offers a streamlined, hosted reinforcement‑learning workflow that lets users customize the NVIDIA Nemotron 3 Nano in minutes. The process establishes a baseline, trains the model o
arXiv:2607.20205v1 Announce Type: cross Abstract: Low-rank adaptation (LoRA) has become a widely used parameter-efficient fine-tuning method for large language models. Since different modules and laye
TL;DR: Everyone says QLoRA past ~13B is dead on a 24GB card. I got the full 32B FLUX.2 dev transformer QLoRA-training resident on the GPU on a 7900 XTX under native ROCm on Windows (no ZLUDA, no CUDA
arXiv:2607.19442v1 Announce Type: cross Abstract: Machine unlearning is commonly evaluated by matching a retrained oracle on trained probes. In a controlled nonce-fact testbed with a matched retrainin
What's Changed mlx update by @dhiltgen in #17332 model/parsers: finalize incomplete GLM tool calls by @dhiltgen in #17250 docs: update retirements by @mxyng in #17289 model: align Laguna with upstream
arXiv:2607.20270v1 Announce Type: new Abstract: Large language models are increasingly evaluated through the values they endorse, but such evaluations presuppose that models can identify the value exp
What’s one of the biggest bottlenecks stopping organizations from scaling their AI initiatives? It isn’t the capabilities of today’s models — it’s their access to business context and semantic meaning
Behind the scenes at Hugging Face headquarter. *literally* Media We're partnering with @huggingface to investigate an unprecedented security incident. Cyber-capable OpenAI models compromised Hugging F
@Engomar_10 @huggingface @alihfadel @Almalki_io We've seen the outpouring of support for ASRs in under-resourced languages, and we hear you. As a start, we've partnered with @HUMAIN to keep working wi
Holy shit wow We're partnering with @huggingface to investigate an unprecedented security incident. Cyber-capable OpenAI models compromised Hugging Face production during a benchmark evaluation. Shari
At Google Cloud Next 2026 in Las Vegas, MLCommons and Google Cloud demonstrated a powerful new capability for trustworthy medical AI - one that protects patient data, model IP, and benchmark integrity
Troubling … OpenAI disclosed it themselves yesterday. Their models (GPT-5.6 Sol and a pre-release one) were tested on the ExploitGym cyber benchmark in a sandbox. They escaped, exploited a zero-day to
feeling nostalgic, my favorite blogs on RL & reward hacking https://www.alexirpan.com/2018/02/14/rl-hard.html https://lilianweng.github.io/posts/2024-11-28-reward-hacking/ TLDR: An openai model, durin
One of the most interesting investigations of my career. We're partnering with @huggingface to investigate an unprecedented security incident. Cyber-capable OpenAI models compromised Hugging Face prod
arXiv:2607.13716v1 Announce Type: new Abstract: Agentic AI systems increasingly act through heterogeneous runtimes: local coding hooks, SDK tools, browser automation, managed-agent traces, API gateway
arXiv:2607.13120v1 Announce Type: cross Abstract: Inferring gene regulatory networks (GRNs) from single-cell transcriptomic data is crucial for biological discovery, yet existing approaches suffer fro
arXiv:2607.13045v1 Announce Type: cross Abstract: Federated Learning (FL) has emerged as a key paradigm for privacy-preserving collaborative model training across distributed and heterogeneous data so
arXiv:2607.13587v1 Announce Type: cross Abstract: Automatic symbolic music analysis has made substantial progress, yet existing systems are typically designed for a single mode of use, such as full-sc
arXiv:2607.13548v1 Announce Type: new Abstract: Identifying root causes in production microservice failures requires reasoning over large-scale, multimodal telemetry spanning metrics, logs, and traces
arXiv:2408.10581v3 Announce Type: replace Abstract: This work introduces a novel and generalizable multi-view Hand Mesh Reconstruction (HMR) model, named POEM, designed for practical use in real-world
arXiv:2607.13941v1 Announce Type: new Abstract: Video aesthetic assessment (VAA) aims to predict how aesthetically pleasing a video is, yet remains far less explored than other visual assessment tasks
arXiv:2607.13189v1 Announce Type: cross Abstract: We present RAGthoven, our system for SemEval-2026 Task 1 (MWAHAHA), Subtask A (multilingual constrained humor generation in English, Spanish, and Chin
arXiv:2607.13348v1 Announce Type: new Abstract: Autonomous racing overtaking requires balancing competitive performance with safety under nonlinear vehicle dynamics and real-time constraints. Model Pr
arXiv:2607.13421v1 Announce Type: cross Abstract: Spatio-Temporal Video Grounding (STVG) aims to retrieve the visual trajectory of a specific object from a video stream as described by a natural langu
arXiv:2607.13597v1 Announce Type: cross Abstract: Vision-Language-Action (VLA) models inherit rich semantic representations from pretrained Vision-Language Models, yet fine-tuning on limited robot dem
arXiv:2607.13047v1 Announce Type: new Abstract: Parameter decomposition (PD) decomposes neural networks into interpretable computational components that faithfully reflect the original network's opera
arXiv:2607.08803v2 Announce Type: replace-cross Abstract: The push toward large language models for biology (BioLM) has created a need for training corpora that can endow models with a genuine underst
arXiv:2604.02668v2 Announce Type: replace-cross Abstract: Large language models (LLMs) often exhibit sycophancy: agreement with user stance even when it conflicts with the model's opinion. While prior
arXiv:2607.13527v1 Announce Type: new Abstract: Recent video generation models (VGMs) have made substantial progress in visual fidelity, yet their ability to follow long, compositional instructions re
arXiv:2602.11619v2 Announce Type: replace Abstract: Running the same LLM agent on identical inputs yields 2.3-4.2 distinct action sequences per 10 runs; this behavioral variance constitutes a training
arXiv:2607.13022v1 Announce Type: cross Abstract: Many nonlinear physical systems exhibit an initial transient phase in which perturbations grow before nonlinear interactions lead to a statistically s
arXiv:2607.12358v1 Announce Type: new Abstract: Video diffusion models produce high-quality generations but remain slow at inference due to their sequential denoising procedure. Caching-based accelera
arXiv:2607.12293v1 Announce Type: new Abstract: Predicting pedestrian crossing intention is a safety-critical task for autonomous driving, yet existing approaches often rely on single-modal inputs or
arXiv:2607.12500v1 Announce Type: cross Abstract: Deep learning models for online handwriting recognition have been shown effective and are increasingly deployed in practical applications. However, th
arXiv:2607.12048v1 Announce Type: cross Abstract: Deploying medical visual question answering (MedVQA) systems in real-world clinical settings requires models that adapt to new clinical tasks without
LLMs handle speech well once you run speech-to-text. They don't hear the rest: a bird outside, a glass breaking two rooms away, a smoke alarm two floors down. I've been working on an experimental open
arXiv:2607.11263v2 Announce Type: replace Abstract: Two competing perspectives on fluid intelligence (gf) measures propose that performance is primarily constrained either by working memory capacity o
arXiv:2607.12466v1 Announce Type: new Abstract: Aligning robot policies with human preferences is essential for deployment to diverse end users. In per-user alignment approach, preference feedback is
arXiv:2607.12934v1 Announce Type: new Abstract: Remote sensing change detection (RSCD) models are prone to catastrophic forgetting when incrementally adapted to new domains. Existing domain-incrementa
arXiv:2607.12544v1 Announce Type: new Abstract: Thermal infrared (TIR) imaging is essential for UAV swarm operations in visually degraded environments. However, tracking tiny UAVs remains challenging
arXiv:2501.05396v3 Announce Type: replace Abstract: Large language models (LLMs) are increasingly used in high-stakes decisions such as hiring and college admissions, making their social bias a critic
arXiv:2606.16847v3 Announce Type: replace-cross Abstract: Diffusion Large Language Models (dLLMs) offer a promising avenue for parallel generation but face a trade-off between decoding speed and quali