All entries

LoRA-Mixer: Coordinate Modular LoRA Experts Through Serial Attention Routing

arXiv:2507.00029v2 Announce Type: replace-cross Abstract: Recent attempts to combine low-rank adaptation (LoRA) with mixture-of-experts (MoE) for multi-task adaptation of Large Language Models (LLMs)

Human

Content type

All Blog X Post Paper YouTube Reddit GitHub

Clear filters

Local Ai

LoREnc: Low-Rank Encryption for Securing Foundation Models and LoRA Adapters

DGX agent

arXiv:2605.13163v1 Announce Type: cross Abstract: Foundation models and low-rank adapters enable efficient on-device generative AI but raise risks such as intellectual property leakage and model recov

local-aiarxiv-cs-cv

14 May 2026

Low-Rank Adapters Initialization via Gradient Surgery for Continual Learning

model-releasesarxiv-cs-lg

arXiv:2605.12752v1 Announce Type: new Abstract: LoRA is widely adopted for continual fine-tuning of Large Language Models due to its parameter efficiency, modularity across tasks, and compatibility wi

LY Corp launches a bid with Bain Capital to buy Kakaku.com, which operates restaurant review and booking site Tabelog, for 4B, challenging EQT's 3.75B offer (Nikkei Asia)

Nikkei Asia: LY Corp launches a bid with Bain Capital to buy Kakaku.com, which operates restaurant review and booking site Tabelog, for 4B, challenging EQT's 3.75B offer — TOKYO — The operator of the

model-releasestechmeme

M2Retinexformer: Multi-Modal Retinexformer for Low-Light Image Enhancement

arXiv:2605.12556v1 Announce Type: new Abstract: Low-light image enhancement is challenging due to complex degradations, including amplified noise, artifacts, and color distortion. While Retinex-based

M3Net: A Macro-to-Meso-to-Micro Clinical-inspired Hierarchical 3D Network for Pulmonary Nodule Classification

arXiv:2605.12570v1 Announce Type: new Abstract: The accurate classification of benign and malignant pulmonary nodules in CT scans is critical for early lung cancer screening, yet remains challenging d

Machine Learning-Driven Multimodal Spectroscopic Liquid Biopsy for Early Multicancer Detection

arXiv:2605.13218v1 Announce Type: new Abstract: Cancer is one of the leading causes of death worldwide, making the development of rapid, minimally invasive, label-free and scalable diagnostic strategi

Macro-Action Based Multi-Agent Instruction Following through Value Cancellation

arXiv:2605.12655v1 Announce Type: new Abstract: Multi-agent reinforcement learning (MARL) in real-world use cases may need to adapt to external natural language instructions that interrupt ongoing beh

safetyarxiv-cs-ai

Hardware

Made a simple utility to share your GPU profile traces. `uvx trace-tuil <local traces> -b <hf bucket name>`

A utility tool was created to streamline sharing GPU profile traces by uploading them to Hugging Face buckets using the command `uvx trace-tuil -b `. This simplifies the process of distributing perfor

hardwareclem-delangue--x

Make-It-Poseable: Feed-forward Latent Posing Model for 3D Characters

arXiv:2512.16767v2 Announce Type: replace Abstract: Posing 3D characters is a fundamental task in computer graphics. However, existing paradigms, ranging from traditional auto-rigging to recent pose-c

Making humans responsible for their AI use seems like an incredibly reasonable way to address problems & opportunities in the use of AI for …

Making humans responsible for their AI use seems like an incredibly reasonable way to address problems & opportunities in the use of AI for academic research, at least in the short term (autonomous sc

agentsethan-mollick--x

MambaPanoptic: A Vision Mamba-based Structured State Space Framework for Panoptic Segmentation

arXiv:2605.12640v1 Announce Type: new Abstract: Panoptic segmentation requires the simultaneous recognition of countable thing instances and amorphous stuff regions, placing joint demands on long-rang

Manipulation Planning for Construction Activities with Repetitive Tasks

arXiv:2605.13754v1 Announce Type: new Abstract: In this paper, we study the problem of manipulation skill acquisition for performing construction activities consisting of repetitive tasks (e.g., build

researcharxiv-cs-ro

Many-Shot CoT-ICL: Making In-Context Learning Truly Learn

arXiv:2605.13511v1 Announce Type: cross Abstract: In-context learning (ICL) adapts large language models (LLMs) to new tasks by conditioning on demonstrations in the prompt without parameter updates.

MAP: A Map-then-Act Paradigm for Long-Horizon Interactive Agent Reasoning

arXiv:2605.13037v1 Announce Type: new Abstract: Current interactive LLM agents rely on goal-conditioned stepwise planning, where environmental understanding is acquired reactively during execution rat

Margin-calibrated Classifier Guidance for Property-driven Synthesis Planning

arXiv:2605.13101v1 Announce Type: cross Abstract: Synthesis planning seeks an efficient sequence of chemical reactions that produce a target molecule. Typically, a pretrained single-step (autoregressi

MARLIN: Multi-Agent Game-Theoretic Reinforcement Learning for Sustainable LLM Inference in Cloud Datacenters

arXiv:2605.13496v1 Announce Type: cross Abstract: Large Language Models (LLMs) have become increasingly prevalent in cloud-based platforms, propelled by the introduction of AI-based consumer and enter

agentsarxiv-cs-lg

MaskPro: Linear-Space Probabilistic Learning for Strict (N:M)-Sparsity on LLMs

arXiv:2506.12876v2 Announce Type: replace Abstract: The rapid scaling of large language models~(LLMs) has made inference efficiency a primary bottleneck in the practical deployment. To address this, s

safetyarxiv-cs-lg

McCast: Memory-Guided Latent Drift Correction for Long-Horizon Precipitation Nowcasting

arXiv:2605.13197v1 Announce Type: cross Abstract: Existing precipitation nowcasting methods typically adopt an autoregressive formulation, where future states are predicted from previous outputs. Howe

Mechanism Plausibility in Generative Agent-Based Modeling

arXiv:2605.12824v1 Announce Type: cross Abstract: Large language models (LLMs) can generate high-level diverse phenomena without explicitly programmed rules. This capability has led to their adoption

Mechanistic Evidence for Spectral Structures in Prior-Data Fitted Networks

model-releasesarxiv-cs-lg

arXiv:2601.21731v2 Announce Type: replace Abstract: Prior-Data Fitted Networks (PFNs) enable amortized Bayesian inference in a single forward pass, yet their internal representations remain opaque. It

MedCore: Boundary-Preserving Medical Core Pruning for MedSAM

model-releasesarxiv-cs-cv

arXiv:2605.13688v1 Announce Type: new Abstract: Medical segmentation foundation models such as SAM and MedSAM provide strong prompt-driven segmentation, but their image encoders are still too large fo

MedOpenClaw and MedFlowBench: Auditing Medical Agents in Full-Study Workflows

model-releasesarxiv-cs-cv

arXiv:2603.24649v2 Announce Type: replace Abstract: Medical imaging benchmarks often evaluate VLMs on pre-selected 2D images, slices, crops, or patches, making evaluation closer to visual recognition.

Meet Kimi Web Bridge - Kimi's browser extension. Agent can now interact with websites like a human: search, scroll, click, type and complete…

model-releaseskimi-moonshot--x

Meet Kimi Web Bridge - Kimi's browser extension. Agent can now interact with websites like a human: search, scroll, click, type and complete tasks. Supports Kimi Code CLI, Claude Code, Cursor, Codex,

Meta observation: DeepSeek is still king of the active-parameter ratio

model-releasessebastian-raschka--x

DeepSeek maintains the highest efficiency in terms of active parameters relative to total model size, outperforming competitors in the ratio of parameters actually used during inference versus total t

Applications

Microsoft and Red Hat are turning a decade-old alliance into enterprise IT’s unified future

Open enterprise hybrid cloud is no longer a buzzword — it’s the answer enterprises are landing on as AI workloads, sovereignty mandates and a decade of accumulated Windows-Linux complexity finally dem

applicationssiliconangle

Microsoft starts canceling Claude Code licenses

model-releasesthe-verge-ai

Microsoft first started opening up access to Claude Code in December, inviting thousands of its own developers to use Anthropic's AI coding tool daily. It was part of an effort to get project managers

Industry

Microsoft’s LinkedIn is about to lay off 5% of its staff

Microsoft Corp.-owned LinkedIn is about to lay off about 5% of its staff, according to report in Reuters today. Speaking on the condition of anonymity, a person familiar with the matter told Reuters t

industrysiliconangle

MIDST Challenge at SaTML 2025: Membership Inference over Diffusion-models-based Synthetic Tabular data

arXiv:2603.19185v2 Announce Type: replace Abstract: Synthetic data is often perceived as a silver-bullet solution to data anonymization and privacy-preserving data publishing. Drawn from generative mo

Applications

MILM: Large Language Models for Multimodal Irregular Time Series with Informative Sampling

arXiv:2605.13711v1 Announce Type: new Abstract: Multimodal irregular time series (MITS) consist of asynchronous and irregularly sampled observations from heterogeneous numerical and textual channels.

applicationsarxiv-cs-lg

Tutorials

Min Generalized Sliced Gromov Wasserstein: A Scalable Path to Gromov Wasserstein

arXiv:2605.13753v1 Announce Type: cross Abstract: We propose min Generalized Sliced Gromov--Wasserstein (min-GSGW), a sliced formulation for the Gromov--Wasserstein (GW) problem using expressive gener

tutorialsarxiv-cs-cv

Min-Max Optimization Requires Exponentially Many Queries

arXiv:2605.13806v1 Announce Type: cross Abstract: We study the query complexity of min-max optimization of a nonconvex-nonconcave function f over [0,1]^d imes [0,1]^d. We show that, given oracle acces

Mind the Gap: How Elicitation Protocols Shape the Stated-Revealed Preference Gap in Language Models

arXiv:2601.21975v2 Announce Type: replace Abstract: Recent work identifies a stated-revealed (SvR) preference gap in language models (LMs): a mismatch between the values models endorse and the choices

MindVLA-U1: VLA Beats VA with Unified Streaming Architecture for Autonomous Driving

model-releasesarxiv-cs-cv

arXiv:2605.12624v1 Announce Type: cross Abstract: Autonomous driving has progressed from modular pipelines toward end-to-end unification, and Vision-Language-Action (VLA) models are a natural extensio

MinT: Managed Infrastructure for Training and Serving Millions of LLMs

arXiv:2605.13779v1 Announce Type: cross Abstract: We present MindLab Toolkit (MinT), a managed infrastructure system for Low-Rank Adaptation (LoRA) post-training and online serving. MinT targets a set

safetyarxiv-cs-ai

Tools

Mist v3 is available across two endpoints: 👉 rime-labs/rime-mist-v3 for English TTS 👉 rime-labs/rime-mist-v3-omni for multilingual TTS acr…

Mist v3 is available across two endpoints: 👉 rime-labs/rime-mist-v3 for English TTS 👉 rime-labs/rime-mist-v3-omni for multilingual TTS across English, Spanish, French, and German 👉 Custom pronunciatio

toolstogether-ai--x

Tools

@mitchellh: 'Programming languages used to be LOCK IN, and they're increasingly not so.'

Mitchell Hashimoto argues that programming languages have historically created vendor lock-in effects but are increasingly losing that characteristic, likely due to standardization, interoperability i

toolssimon-willison--x

Tools

Mitchell's post here reminded me of a similar conversation I had recently about how cheap it can be to port native mobile apps to React Nati…

Mitchell's post here reminded me of a similar conversation I had recently about how cheap it can be to port native mobile apps to React Native using coding agents... and then port them back again late

toolssimon-willison--x

Mix, Don't Tune: Bilingual Pre-Training Outperforms Hyperparameter Search in Data-Constrained Settings

arXiv:2605.13225v1 Announce Type: new Abstract: For most languages of the world, language model pre-training operates in a data-constrained regime where models must repeat their training data many tim

Mixed neural posterior estimation for simulators with discrete and continuous parameters

model-releasesarxiv-cs-lg

arXiv:2605.13551v1 Announce Type: new Abstract: Neural Posterior Estimation (NPE) enables rapid parameter inference for complex simulators with intractable likelihoods. NPE trains an inference network

MLGIB: Multi-Label Graph Information Bottleneck for Expressive and Robust Message Passing

arXiv:2605.13126v1 Announce Type: cross Abstract: Graph Neural Networks (GNNs) suffer from over-squashing in deep message passing, where information from exponentially growing neighborhoods is compres

MMCL-Bench: Multimodal Context Learning from Visual Rules, Procedures, and Evidence

arXiv:2605.12703v1 Announce Type: cross Abstract: We introduce MMCL-Bench, a benchmark for multimodal context learning: learning task-local rules, procedures, and empirical patterns from visual or mix

MMSkills: Towards Multimodal Skills for General Visual Agents

arXiv:2605.13527v1 Announce Type: new Abstract: Reusable skills have become a core substrate for improving agent capabilities, yet most existing skill packages encode reusable behavior primarily as te

MobiBench: Multi-Branch, Modular Benchmark for Mobile GUI Agents

arXiv:2512.12634v3 Announce Type: replace Abstract: Mobile GUI Agents, AI agents capable of interacting with mobile applications on behalf of users, have the potential to transform human computer inte

MoCCA: A Movable Circle Probability of Collision Approximation

arXiv:2605.13125v1 Announce Type: new Abstract: In automated driving, crash mitigation is crucial to ensure passenger safety. Accurate avoidance requires precise knowledge of the object's position and

safetyarxiv-cs-ro

Model. Harness. Context. The 3 main components of agents. As you build more agents, context increasingly lives AGENTS.md, skills, policies, …

Model. Harness. Context. The 3 main components of agents. As you build more agents, context increasingly lives AGENTS.md, skills, policies, examples, + generated research files. Context needs its own

agentsharrison-chase--x

Tutorials

Modeling Heterophily in Multiplex Graphs: An Adaptive Approach for Node Classification

arXiv:2605.12699v1 Announce Type: cross Abstract: Existing multiplex graph models often assume homophily, where connected nodes tend to belong to the same class or share similar attributes. Consequent

tutorialsarxiv-cs-ai

Moltbook Moderation: Uncovering Hidden Intent Through Multi-Turn Dialogue

arXiv:2605.12856v1 Announce Type: new Abstract: The emergence of multi-agent systems introduces novel moderation challenges that extend beyond content filtering. Agents with {em malicious intent} may