research

Mechanistic Independence: A Principle for Identifiable Disentangled Representations

arXiv:2509.22196v2 Announce Type: replace Abstract: Disentangled representations seek to recover latent factors of variation underlying observed data, yet their identifiability is still not fully unde

Research

MedThink: Enhancing Diagnostic Accuracy in Small Models via Teacher-Guided Reasoning Correction

DGX agent

arXiv:2605.08094v1 Announce Type: cross Abstract: Accurate clinical diagnosis requires extensive domain knowledge and complex clinical reasoning capabilities. Although large language models (LLMs) hol

researcharxiv-cs-ai

12 May 2026

Mela: Test-Time Memory Consolidation based on Transformation Hypothesis

arXiv:2605.10537v1 Announce Type: new Abstract: Memory consolidation, the process by which transient experiences are transformed into stable, structured representations, is a foundational organizing p

researcharxiv-cs-cl

Memorize Theorems, Not Instances: Probing SFT Generalization through Mathematical Reasoning

arXiv:2605.09270v1 Announce Type: cross Abstract: Supervised Fine-Tuning (SFT) is widely used for task-specific adaptation, yet recent work shows it systematically undermines reasoning generalization.

MeshFIM: Local Low-Poly Mesh Editing via Fill-in-the-Middle Autoregressive Generation

arXiv:2605.08744v1 Announce Type: cross Abstract: Autoregressive (AR) models can generate high-quality low-poly meshes from point clouds, but they still operate in an all-or-nothing manner: when a loc

Metacognitive Behavioral Tuning of Large Language Models for Multi-Hop Question Answering

arXiv:2602.22508v2 Announce Type: replace Abstract: Large Language Models (LLMs) often produce incorrect answers on multi-hop question answering even when the reasoning trace already contains a correc

METBRA25Y: Brazil Surface Meteorology Archive with Harmonized Variables and Quality Control

arXiv:2605.08701v1 Announce Type: new Abstract: This data paper describes METBRA25Y, a harmonized archive of hourly surface meteorological observations from Brazil derived from public historical recor

Micro-Defects Expose Macro-Fakes: Detecting AI-Generated Images via Local Distributional Shifts

arXiv:2605.09296v1 Announce Type: cross Abstract: Recent generative models can produce images that appear highly realistic, raising challenges in distinguishing real and AI-generated images. Yet exist

MicroDiffuse3D: A Foundation Model for 3D Microscopy Imaging Restoration

arXiv:2605.08566v1 Announce Type: new Abstract: Chemical imaging enables label-free visualization of cells, tissues and living systems while providing direct biochemical information that is difficult

MicroViTv2: Beyond the FLOPS for Edge Energy-Friendly Vision Transformers

arXiv:2605.10148v1 Announce Type: new Abstract: The Vision Transformer (ViT) achieves remarkable accuracy across visual tasks but remains computationally expensive for edge deployment. This paper pres

Mind the Gap No More: Achieving Zero-Gap Multimodal Integration via One Tokenizer

arXiv:2602.12286v2 Announce Type: replace-cross Abstract: A central challenge in developing Multimodal Large Language Models (MLLMs) is effectively integrating heterogeneous inputs into a cohesive rea

researcharxiv-cs-cl

MineEvolve: Self-Evolution with Accumulated Knowledge for Long-Horizon Embodied Minecraft Agents

arXiv:2603.13131v3 Announce Type: replace Abstract: Long-horizon embodied intelligence requires agents to improve through interaction, not merely to execute plans generated from static goals. A centra

Minimal Filling Architectures of Polynomial Neural Networks: Counterexamples, Frontier Search, and Defects

arXiv:2605.09609v1 Announce Type: new Abstract: We provide a counterexample to the minimal unimodal conjecture for polynomial neural networks (PNNs) with power activation functions. Fixing the input a

Mistake-Bounded Language Generation

arXiv:2605.10809v1 Announce Type: new Abstract: We investigate the learning task of language generation in the limit, but shift focus from the traditional time-of-last-mistake metric of a generator's

Mitigating Membership Inference in Intermediate Representations with Differentially Private Training

arXiv:2602.22611v2 Announce Type: replace Abstract: In Embedding-as-an-Interface (EaaI) settings, pre-trained models are queried for Intermediate Representations (IRs). The distributional properties o

Mitigating Watermark Forgery in Generative Models via Randomized Key Selection

arXiv:2507.07871v4 Announce Type: replace-cross Abstract: Watermarking enables GenAI providers to verify whether content was generated by their models. A watermark is a hidden signal in the content, w

Mixture of Experts for Recognizing Depression from Interview and Reading Tasks

arXiv:2502.20213v2 Announce Type: replace Abstract: Depression is a mental disorder and can cause a variety of symptoms, including psychological, physical, and social. Speech has been proved an object

Mixture of Layers with Hybrid Attention

arXiv:2605.09516v1 Announce Type: cross Abstract: Standard Mixture-of-Experts (MoE) transformers route tokens to expert subnetworks within each layer, but the layer structure itself remains monolithic

ML-CLIPSim: Multi-Layer CLIP Similarity for Machine-Oriented Image Quality

arXiv:2605.09479v1 Announce Type: cross Abstract: We study full-reference image quality assessment from a machine-centric perspective, where images are evaluated by how well they preserve information

MOBIUS: A Multi-Modal Bipedal Robot that can Walk, Crawl, Climb, and Roll

arXiv:2511.01774v3 Announce Type: replace Abstract: This paper presents the MOBIUS platform, a bipedal robot capable of walking, crawling, climbing, and rolling. MOBIUS features four limbs, two 6-DoF

researcharxiv-cs-ro

Model-based Dynamic 3D MRI Reconstructions using Neural Fields and Tensor Product Expansions

arXiv:2605.08275v1 Announce Type: cross Abstract: Conventional MRI reconstruction methods treat images and coil sensitivities as discrete objects, leading to high memory demands and limited structural

Model Merging Scaling Laws in Large Language Models

arXiv:2509.24244v4 Announce Type: replace Abstract: We study empirical scaling laws for language model merging measured by cross-entropy. Despite its wide practical use, merging lacks a quantitative r

Model-Reference Adaptive Flight Control of the 95-mg Bee++

arXiv:2605.08525v1 Announce Type: new Abstract: We introduce a model-reference adaptive control (MRAC) architecture for high-performance positional tracking of the Bee++, a 95-mg insect-scale flapping

researcharxiv-cs-ro

Modeling Atomic Conformational Ensembles of Proteins via Test-Time Supervision of Boltz-2 on Cryo-EM Density Maps

arXiv:2605.09832v1 Announce Type: new Abstract: Knowledge of a protein's atomic conformational ensemble is critical to determining its function, yet state-of-the-art ensemble prediction models are lim

MoPO: Incorporating Motion Prior for Occluded Human Mesh Recovery

arXiv:2605.09856v1 Announce Type: cross Abstract: Although recent studies have made remarkable progress in human mesh recovery, they still exhibit limited robustness to occlusions and often produce in

Multi-Armed Bandits With Best-Action Queries

arXiv:2605.08287v1 Announce Type: cross Abstract: We study multi-armed bandits (MABs) augmented with best-action queries, in which the learner may additionally query an oracle that reveals the best ar

Multi-Level Graph Attention Network Contrastive Learning for Knowledge-Aware Recommendation

arXiv:2605.08499v1 Announce Type: cross Abstract: In recent years, the use of edge information provided by knowledge graphs together with the advantages of higher-order connectivity in graph neural ne

MultiAnimate: Pose-Guided Image Animation Made Extensible

arXiv:2602.21581v2 Announce Type: replace Abstract: Pose-guided human image animation aims to synthesize realistic videos of a reference character driven by a sequence of poses. While diffusion-based

Multifidelity Gaussian process regression for solving nonlinear partial differential equations

arXiv:2605.10383v1 Announce Type: cross Abstract: Solving nonlinear partial differential equations (PDEs) using kernel methods offers a compelling alternative to traditional numerical solvers. However

MultiMedVision: Multi-Modal Medical Vision Framework

arXiv:2605.09151v1 Announce Type: new Abstract: Multi-modal medical imaging enables comprehensive diagnostics, yet current foundation models process 2D (e.g. X-ray) and 3D (e.g. CT) data with separate

Multimodal Emotion Recognition via Causal-Diffusion Bridge (Affect-Diff)

arXiv:2605.08252v1 Announce Type: new Abstract: Multimodal emotion recognition on CMU-MOSEI faces an extreme imbalance as Happy accounts for 65.9% of samples while three Ekman categories collectively

Muon Does Not Converge on Convex Lipschitz Functions

arXiv:2605.08980v1 Announce Type: new Abstract: Muon and its variants have shown strong empirical performance in a variety of deep learning tasks. Existing convergence analyses of Muon rely on smoothn

MuonEq: Balancing Before Orthogonalization with Lightweight Equilibration

arXiv:2603.28254v2 Announce Type: replace Abstract: Orthogonalized-update optimizers such as Muon improve training of matrix-valued parameters, but existing extensions typically either rescale updates

Muown: Row-Norm Control for Muon Optimization

arXiv:2605.10797v1 Announce Type: new Abstract: Muon has emerged as a strong competitor to AdamW for language model pre-training, yet its behavior at scale is sensitive to weight decay. Recent work ha

NaiAD: Initiate Data-Driven Research for LLM Advertising

arXiv:2605.09918v1 Announce Type: cross Abstract: Reconciling platform revenue with user experience in LLM advertising motivates a data-centric foundation. We introduce NaiAD, the first comprehensive

Narrative Landscape: Mapping Narrative Dispositions Across LLMs

arXiv:2605.08742v1 Announce Type: cross Abstract: This study proposes a quantitative framework for profiling LLM dispositions as stable, model-specific regularities in output under repeated, controlle

NCO: A Versatile Plug-in for Handling Negative Constraints in Decoding

arXiv:2605.10065v1 Announce Type: cross Abstract: Controlling Large Language Models (LLMs) to prevent the generation of undesirable content, such as profanity and personally identifiable information (

Near-Optimal Last-Iterate Convergence for Zero-Sum Games with Bandit Feedback and Opponent Actions

arXiv:2605.09363v1 Announce Type: new Abstract: Last-iterate convergence of learning dynamics in games has attracted significant recent attention. In two-player zero-sum games with bandit feedback, wh

Nearly-Optimal Algorithm for Adversarial Kernelized Bandits

arXiv:2605.10299v1 Announce Type: new Abstract: This paper studies kernelized bandits (also known as Gaussian process bandits) in an adversarial environment, where the reward functions in a known repr

Network-Efficient World Model Token Streaming

arXiv:2605.09886v1 Announce Type: new Abstract: Generative driving world models rely on compact latent state representations that must be efficiently transmitted and synchronized across distributed co

researcharxiv-cs-ro

Neurally-plausible radial basis kernels using distributed Fourier embeddings

arXiv:2605.08458v1 Announce Type: new Abstract: Coherent, continuous spatial representations are critical for synthesizing physical and perceptual phenomena into a single representational space. Radia

Neuromorphic Monocular Depth Estimation with Uncertainty Modeling

arXiv:2605.10675v1 Announce Type: new Abstract: Event cameras offer distinct advantages over conventional frame-based sensors, including microsecond-level temporal resolution, high dynamic range, and

Neuroprobe: Evaluating Intracranial Brain Responses to Naturalistic Stimuli

arXiv:2509.21671v2 Announce Type: replace Abstract: High-resolution neural datasets enable foundation models for the next generation of brain-computer interfaces and neurological treatments. The commu

Neuroscience-Inspired Analyses of Visual Interestingness in Multimodal Transformers

arXiv:2605.08188v1 Announce Type: cross Abstract: Human attention is the gateway to conscious perception, memory and decision-making. However, its role in modern transformer models remains largely une

Nix and Fix: Targeting 1000x Compression of 3D Gaussian Splatting with Diffusion Models

arXiv:2602.04549v2 Announce Type: replace Abstract: 3D Gaussian Splatting (3DGS) revolutionized novel view rendering. Instead of inferring from dense spatial points, as implicit representations do, 3D

No Mean Feat: Simple, Strong Baselines for Context Compression

arXiv:2510.20797v2 Announce Type: replace-cross Abstract: Context compression reduces Transformer inference costs by replacing lengthy inputs with shorter pre-computed representations. It carries sign

NoiseRater: Meta-Learned Noise Valuation for Diffusion Model Training

arXiv:2605.08144v1 Announce Type: cross Abstract: Diffusion models have achieved remarkable success across a wide range of generative tasks, yet their training paradigm largely treats injected noise a

NoisyCoconut: Counterfactual Consensus via Latent Space Reasoning

arXiv:2605.08221v1 Announce Type: cross Abstract: This paper presents NoisyCoconut, a novel inference-time method that enhances large language model (LLM) reliability by manipulating internal represen