Search: “models”

Thaka at KSAA-2026 Task 2: Regularized Fine-Tuning for Arabic Speech Diacritization

arXiv:2605.25928v1 Announce Type: new Abstract: We describe the winning system for Task 2 of the KSAA-2026 Shared Task on Arabic Speech Dictation with Automatic Diacritization. The task requires produ

Research

.@tinkerapi by TML is a highly underrated product. It should be part of the “starter post training” infra for any newco

DGX agent

TinkerAPI, developed by TML, is a post-training infrastructure tool that Soumith Chintala recommends as an essential component for new AI companies during their startup phase. The tool appears to addr

researchsoumith-chintala--x

26 May 2026

Tool-Call Dependency Structure is Linearly Decodable in LLM Agent Residual Streams

arXiv:2605.25310v1 Announce Type: new Abstract: Tool-using LLM agents produce trajectories whose calls form a directed dependency graph: earlier tool outputs supply arguments to later calls. Whether t

agentsarxiv-cs-cl

Towards a Universal Causal Reasoner

arXiv:2605.24873v1 Announce Type: cross Abstract: Despite the importance of causal reasoning, training LLMs to reason causally remains underexplored. Existing data efforts mostly focus on benchmarking

applicationsarxiv-cs-ai

Toxicity in Twitch Chats: An LLM-Based Analysis Across Gaming Communities

arXiv:2605.24000v1 Announce Type: new Abstract: Toxicity in online gaming communities remains a persistent challenge, manifesting across genres, platforms, and player interactions. While much research

Trust-Aware Joint Feature-Prediction Discrepancy for Robust Domain Adaptation

arXiv:2605.25119v1 Announce Type: cross Abstract: Domain adaptation aims to mitigate performance degradation caused by distribution shifts between a labeled source domain and an unlabeled or sparsely

Uncertainty-DTW for Sequences and Visual Tokens

arXiv:2605.25110v1 Announce Type: cross Abstract: Aligning structured data is a fundamental problem in computer vision and machine learning, underlying tasks such as time series analysis, human action

Unveil: Unified Visual-Textual Integration and Distillation for Multi-modal Document Retrieval

arXiv:2605.24530v1 Announce Type: new Abstract: Document retrieval in real-world scenarios faces significant challenges due to diverse document formats and modalities. Traditional text-based approache

applicationsarxiv-cs-cl

What Gets Cited: Competitive GEO in AI Answer Engines

arXiv:2605.25517v1 Announce Type: new Abstract: AI answer engines generate answers from retrieved pages but cite only a few sources. This makes visibility depend not just on ranking, but on being cite

Advanced AI Service Provisioning in O-RAN through LLM Engine Integration

arXiv:2605.23809v1 Announce Type: cross Abstract: The Open Radio Access Network (O-RAN) architecture allows AI to be embedded directly into the RAN through modular xApps and rApps, yet creating these

Approximate Machine Unlearning through Manifold Representation Forgetting Guided by Self Mode Connectivity

arXiv:2605.22871v1 Announce Type: cross Abstract: Machine unlearning is a fundamental mechanism that enforces the right to be forgotten. Existing unlearning studies that rely on label manipulation or

local-aiarxiv-cs-ai

b9315

local-aillama-cpp-releases

Build b9315 is a release version of llama.cpp, a C/C++ implementation for LLM inference. As an intermediate build in the llama.cpp development sequence, b9315 likely includes bug fixes, performance im

Building a privacy-preserving Federated Recommender system for mobile devices

arXiv:2605.22924v1 Announce Type: new Abstract: Serving personalized content on mobile devices has traditionally required pooling sensitive user data on centralized servers, a practice increasingly at

local-aiarxiv-cs-lg

Tutorials

CALAD: Channel-Aware contrastive Learning for multivariate time series Anomaly Detection

arXiv:2605.23139v1 Announce Type: cross Abstract: Multivariate time series anomaly detection has become increasingly important in real-world applications, where labeled data are often scarce. Many exi

tutorialsarxiv-cs-ai

Cascaded Transfer: Learning Many Tasks under Budget Constraints

arXiv:2601.21513v2 Announce Type: replace Abstract: In distributed applications, such as energy demand forecasting at the substation level or federated learning, a large number of related tasks must b

Commutator-Induced Uncertainty in VAEs

arXiv:2605.23449v1 Announce Type: cross Abstract: Variational autoencoders (VAEs) often struggle to represent non-commutative structure in learned latent spaces. Symmetry-aware VAEs commonly address t

researcharxiv-cs-cv

CoMoGen: COntrollable MOtion Dynamics and Interactions with Mask-Guided Video GENeration

arXiv:2605.22996v1 Announce Type: new Abstract: We present CoMoGen, a controllable video generation framework that generates realistic interactive dynamics from a single binary mask sequence condition

researcharxiv-cs-cv

CoReVAD: A Contextual Reasoning Framework for Training-Free Video Anomaly Detection

arXiv:2605.23116v1 Announce Type: cross Abstract: Existing Video Anomaly Detection (VAD) methods typically rely on task-specific training, leading to strong domain dependency and high training costs.

DCGAN inference on a microcontroller: 12.6M parameters, 512KB SRAM, 26-second generation, pure C [P]

researchr-machinelearning

This post describes implementing DCGAN (Deep Convolutional Generative Adversarial Network) inference on resource-constrained microcontroller hardware, achieving image generation with a 12.6 million pa

DELICATE: Diachronic Entity LInking using Classes And Temporal Evidence

arXiv:2511.10404v2 Announce Type: replace Abstract: In spite of the remarkable advancements in the field of Natural Language Processing, the task of Entity Linking (EL) remains challenging in the fiel

Enhancing 3D Semantic Scene Completion with a Refinement Module

arXiv:2512.18363v2 Announce Type: replace Abstract: We propose ESSC-RM, a plug-and-play Enhancing framework for Semantic Scene Completion with a Refinement Module, which can be seamlessly integrated i

researcharxiv-cs-cv

Eye Gaze-Informed and Context-Aware Pedestrian Trajectory Prediction in Shared Spaces with Automated Shuttles: A Virtual Reality Study

arXiv:2603.19812v2 Announce Type: replace Abstract: To address this gap, we conduct a Virtual Reality experiment in which pedestrians interact with automated shuttles under varying approach angles (45

Forget by Uncertainty: Orthogonal Entropy Unlearning for Quantized Neural Networks

arXiv:2602.00567v2 Announce Type: replace Abstract: The deployment of quantized neural networks on edge devices, combined with privacy regulations like GDPR, creates an urgent need for machine unlearn

From Activation to Causality: Discovery of Causal Visual Representations in the Human Brain

arXiv:2605.23895v1 Announce Type: new Abstract: Identifying which brain regions represent a visual concept in the human brain is a central challenge in neuroscience. Existing approaches have localized

local-aiarxiv-cs-cv

FusionSense: Tri-Stage Near-Sensor Learning for Runtime-Adaptive Multimodal Edge Intelligence

arXiv:2605.22868v1 Announce Type: new Abstract: Autonomous systems and smart-industry deployments increasingly split computation across near-sensor, edge, and cloud resources, where tight energy, late

agentsarxiv-cs-lg

GenRecon: Bridging Generative Priors for Multi-View 3D Scene Reconstruction

arXiv:2605.23888v1 Announce Type: new Abstract: We introduce a new approach to high-fidelity 3D scene reconstruction from multi-view RGB images that tightly couples reconstruction with a strong genera

local-aiarxiv-cs-cv

GFSR: Geometric Fidelity and Spatial Refinement for Reliable Lane Detection

arXiv:2605.23327v1 Announce Type: new Abstract: Lane detection stands as a crucial perception task in autonomous driving and advanced driver assistance systems. However, existing methods still degrade

agentsarxiv-cs-cv

Graph-based Complexity Forecasts in UK En Route Airspace Using Relevant Aircraft Interactions

arXiv:2605.23696v1 Announce Type: new Abstract: Effectively managing Air Traffic Control Officer (ATCO) workload is crucial in maintaining operational safety. Group supervisors use tools that estimate

safetyarxiv-cs-lg

How Many Training Samples Are Needed for the Inverse Kinematics Solutions by Artificial Neural Networks

arXiv:2605.23583v1 Announce Type: cross Abstract: Inverse Kinematics (IK) plays a critical role in robotic motion planning and control. The IK solutions of a robot manipulator could be done by convent

applicationsarxiv-cs-lg

Infra-Bayesian Reinforcement Learning Agents Outperform Classical RL For Worst-Case Robustness

arXiv:2605.23146v1 Announce Type: cross Abstract: Classical reinforcement learning assumes the agent interacts with a fixed environment whose behavior does not depend on the agent's policy. This assum

Learning Individual Dynamics from Sparse Cross-Sectional Snapshots

arXiv:2605.23470v1 Announce Type: cross Abstract: Predicting how a dynamical unit evolves over time - how an individual ages, an epidemic spreads, or a physical system degrades - typically requires de

applicationsarxiv-cs-ai

Learning Kernel-Based MDPs from Episodic Preferential Feedback

arXiv:2605.23650v1 Announce Type: cross Abstract: Human feedback often arrives as preferences rather than calibrated numeric rewards, motivating reinforcement learning from preferential feedback, also

safetyarxiv-cs-lg

Tutorials

Less Effort, Shorter Proofs: Reinforcement Learning for Security Protocol Analysis in Tamarin

arXiv:2605.23643v1 Announce Type: cross Abstract: Tools like Tamarin and ProVerif have achieved notable success in analyzing and verifying complex real-world protocols such as EMV, 5G, and WPA2, even

tutorialsarxiv-cs-lg

MapGCLR: Geospatial Contrastive Learning of Representations for Online Vectorized HD Map Construction

arXiv:2603.10688v2 Announce Type: replace-cross Abstract: Autonomous vehicles rely on map information to understand the world around them. However, the creation and maintenance of offline high-definit

agentsarxiv-cs-cv

Naturalistic measure of social norms alignment

arXiv:2605.23420v1 Announce Type: new Abstract: Social norms reflect shared expectations on acceptable behavior. Measuring social norms alignment remains challenging, with existing approaches typicall

safetyarxiv-cs-cl

Not Too Generative, Not Too Discriminative: The Human Alignment Sweet Spot

arXiv:2605.23819v1 Announce Type: cross Abstract: A central question in computational vision is whether human-like visual representations are better explained by discriminative or generative learning.

Tools

Notes on Pope Leo XIV's encyclical on AI

Dropped this morning by the Vatican: Magnifica Humanitas of His Holiness Pope Leo XIV on Safeguarding the Human Person in the Time of Artificial Intelligence. This is a very interesting document. It's

toolssimon-willison

On the Infinite Width and Depth Limits of Predictive Coding Networks

arXiv:2602.07697v2 Announce Type: replace-cross Abstract: Predictive coding (PC) is a biologically plausible alternative to standard backpropagation (BP) that minimises an energy function with respect

researcharxiv-cs-ai

PaP-NF: Probabilistic Long-Term Time Series Forecasting via Prefix-as-Prompt Reprogramming and Normalizing Flows

arXiv:2605.23219v1 Announce Type: cross Abstract: Time series forecasting plays a central role in many real-world applications and has been extensively studied. Most existing approaches rely on determ

applicationsarxiv-cs-ai

PathNavigate: A Training-Free Pathology Agent with Surprise-Guided Scan and Shared Slide Memory for Whole-Slide Image VQA

arXiv:2605.23559v1 Announce Type: cross Abstract: Whole-slide image visual question answering (WSI-VQA) frames pathology as an extreme-context search problem: to answer a free-form clinical query, a s

local-aiarxiv-cs-ai

RAS: Reflection-Augmented Scaling with In-Context Learning for Executable Cypher Query Generation

arXiv:2605.22937v1 Announce Type: new Abstract: Inference-time scaling can reduce errors in structured query generation, but methods to allocate the compute for query code generation remains underexpl

Tutorials

RiGS: Rigid-aware 4D Gaussian Splatting from a Single Monocular Video

arXiv:2605.23672v1 Announce Type: new Abstract: Reconstructing dynamic 3D scenes from monocular videos is a fundamental yet highly challenging task, as real-world motions often involve both long-term

tutorialsarxiv-cs-cv

Robotic Strawberry Harvesting with Robust Vision and Deep Reinforcement Learning based Sim-to-Real Control

arXiv:2605.23863v1 Announce Type: new Abstract: This study presents a closed-loop robotic strawberry harvesting system that combines a robust vision module, simulation-trained deep reinforcement learn

safetyarxiv-cs-ro

Selective Ambulance Dispatch Under Contextual Travel-Time Uncertainty

arXiv:2605.23378v1 Announce Type: cross Abstract: Ambulance response is time-critical in out-of-hospital cardiac arrest (OHCA), where dispatchers must balance timely arrivals with limited fleet capaci

Sparse Compositional Flow Matching by geometric assembly from motion primitives

arXiv:2605.23341v1 Announce Type: cross Abstract: Embodied trajectories, such as the executable motion sequences of robotic manipulators, underwater vehicles, and mobile robots, are a fundamental outp

researcharxiv-cs-ai

Spectral-inspired Operator Learning with Limited Data and Unknown Physics

arXiv:2505.21573v3 Announce Type: replace-cross Abstract: Learning PDE dynamics from limited data with unknown physics is challenging. Existing neural PDE solvers either require large datasets or rely

local-aiarxiv-cs-ai

Staging by the Book: Automatic Sleep Stage Classification Using Scoring Rules

arXiv:2605.22859v1 Announce Type: cross Abstract: Automated sleep staging is commonly approached as a supervised machine learning problem, with deep learning methods dominating recent research. While

researcharxiv-cs-ai

Strong Teacher Not Needed? On Distillation in LLM Pretraining

arXiv:2605.23857v1 Announce Type: cross Abstract: Knowledge distillation generally assumes a strong-to-weak relationship where stronger teachers yield better students. In this work, we examine this as