Search: “models”

Security in the Fine-Tuning Lifecycle of Large Language Models: Threats, Defenses,Evaluation, and Future Directions

arXiv:2605.25073v1 Announce Type: cross Abstract: Background: Fine-tuning is central to adapting pre-trained Large Language Models (LLMs) to downstream tasks, but its reliance on training data, parame

26 May 2026

All Blog X Post Paper YouTube Reddit GitHub

Clear filters

Agents

STORM: Internalized Modeling for Spatial-Temporal Reasoning in Video-Language Models

DGX agent

arXiv:2605.26014v1 Announce Type: cross Abstract: Many video reasoning tasks require tracking motion, temporal order, and evolving visual states across frames. Existing methods built on large vision-l

agentsarxiv-cs-cl

26 May 2026

I-SAFE: Wasserstein Coherence Metrics for Structural Auditing of Scientific AI Models

arXiv:2605.21731v1 Announce Type: new Abstract: Deep learning models are increasingly used in scientific prediction tasks where strong benchmark performance is often interpreted as evidence of scienti

23 May 2026

Maestro: Reinforcement Learning to Orchestrate Hierarchical Model-Skill Ensembles

arXiv:2605.22177v1 Announce Type: cross Abstract: The proliferation of large language models (LLMs) and modular skills has endowed autonomous agents with increasingly powerful capabilities. Existing f

22 May 2026

Pre-VLA: Preemptive Runtime Verification for Reliable Vision-Language-Action and World-Model Rollouts

arXiv:2605.22446v1 Announce Type: new Abstract: While large vision-language-action (VLA) models and generative world models (WM) have advanced long-horizon embodied intelligence, their practical deplo

22 May 2026

Understanding Data Temporality Impact on Large Language Models Pre-training

arXiv:2605.22769v1 Announce Type: new Abstract: Large language models (LLMs) are typically trained on shuffled corpora, yielding models whose knowledge is frozen at train time and whose temporal groun

22 May 2026

Agentic Physical AI toward a Domain-Specific Foundation Model for Nuclear Reactor Control

arXiv:2512.23292v3 Announce Type: replace-cross Abstract: The prevailing paradigm in AI for physical systems (scaling general-purpose foundation models toward universal multimodal reasoning) confronts

Calibration vs Decision Making: Revisiting the Reliability Paradox in Unlearned Language Models

arXiv:2605.20915v1 Announce Type: new Abstract: Machine unlearning aims to remove the influence of specific training data from a model while preserving reliable behavior on the remaining data, making

Local Ai

Diffusion Models Memorize in Training -- and Generalize in Inference

arXiv:2603.13419v2 Announce Type: replace Abstract: Diffusion models generalize well in practice. However, an optimal diffusion model fully memorizes the training data and therefore fails to generaliz

local-aiarxiv-cs-lg

DiMextsuperscript{3}: Bridging Multilingual and Multimodal Models via Direction- and Magnitude-Aware Merging

arXiv:2605.12960v2 Announce Type: replace Abstract: Towards more general and human-like intelligence, large language models should seamlessly integrate both multilingual and multimodal capabilities; h

Improving Quantized Model Performance in Qualitative Analysis with Multi-Pass Prompt Verification

arXiv:2605.20193v1 Announce Type: new Abstract: Quantized Large Language Models (LLMs) are used more often in qualitative analysis because they run fast and need fewer computing resources. This study

EVA-0: Test-Time Model Evolution with Only Two Forward Passes per Sample

arXiv:2605.18867v1 Announce Type: cross Abstract: Test-time model evolution offers a promising way for deployed models to improve from unlabeled test-time experience, yet most existing methods depend

HalluWorld: A Controlled Benchmark for Hallucination via Reference World Models

arXiv:2605.19341v1 Announce Type: cross Abstract: Hallucination remains a central failure mode of large language models, but existing benchmarks operationalize it inconsistently across summarization,

Lying Is Just a Phase: The Hidden Alignment Transition in Language Model Scaling

arXiv:2605.18838v1 Announce Type: cross Abstract: Scaling laws predict loss from compute but not how capabilities interact. We measure the coupling between reasoning and truthfulness across 63 base mo

Rebalancing Reference Frame Dominance to Improve Motion in Image-to-Video Models

arXiv:2605.19398v1 Announce Type: cross Abstract: Image-to-video models often generate videos that remain overly static, compared to text-to-video models. While prior approaches mitigate this issue by

Unified Deployment-Aware Evaluation of Open Reasoning Language Models

arXiv:2604.07035v2 Announce Type: replace Abstract: Open reasoning language models are often compared under mixed sample sizes, partially standardized prompts, and accuracy-centered summaries, which m

Safety

AutoRubric-T2I: Robust Rule-Based Reward Model for Text-to-Image Alignment

arXiv:2605.17602v1 Announce Type: new Abstract: Aligning Text-to-Image (T2I) generation models with human preferences increasingly relies on image reward models that score or rank generated images acc

safetyarxiv-cs-ai

Generalization or Memorization? Brittleness Testing for Chess-Trained Language Models

arXiv:2605.17565v1 Announce Type: new Abstract: Recent work has fine-tuned language models on chess data and reported high benchmark scores as evidence that the resulting models can understand the rul

HydroAgent: Closing the Gap Between Frontier LLMs and Human Experts in Hydrologic Model Calibration via Simulator-Grounded RL

arXiv:2605.17792v1 Announce Type: new Abstract: Calibrating distributed hydrologic models is a critical bottleneck across operational water resources management - streamflow prediction, reservoir oper

Internalizing Tool Knowledge in Small Language Models via QLoRA Fine-Tuning

arXiv:2605.17774v1 Announce Type: new Abstract: Large language models are increasingly used as planning components in agentic systems, but current tool-use pipelines often require full tool schemas to

KVCapsule: Efficient Sequential KV Cache Compression for Vision-Language Models with Asymmetric Redundancy

arXiv:2605.16439v1 Announce Type: cross Abstract: Vision-Language Models (VLMs) have emerged as a critical and fast-growing extension of Large Language Models (LLMs) that enable multimodal reasoning t

Language-Switching Triggers Take a Latent Detour Through Language Models

arXiv:2605.18646v1 Announce Type: new Abstract: Backdoor attacks on language models pose a growing security concern, yet the internal mechanisms by which a trigger sequence hijacks model computations

SIPO: Stabilized and Improved Preference Optimization for Aligning Diffusion Models

arXiv:2505.21893v3 Announce Type: replace-cross Abstract: Preference learning has garnered extensive attention as an effective technique for aligning diffusion models with human preferences in visual

The Range Shrinks, the Threat Remains: Re-evaluating LLM Package Hallucinations on the 2026 Frontier-Model Cohort

arXiv:2605.17062v1 Announce Type: cross Abstract: Spracklen et al. (USENIX Security '25) showed that code-generating large language models hallucinate package names that do not exist on PyPI or npm at

Agents

Xiaomi EV World Model: A Joint World Model Integrating Reconstruction and Generation for Autonomous Driving

arXiv:2605.18137v1 Announce Type: new Abstract: This report presents a unified technical system addressing the two core capabilities of world models for autonomous driving: world representation and wo

agentsarxiv-cs-cv

GiLT: Augmenting Transformer Language Models with Dependency Graphs

arXiv:2605.15562v1 Announce Type: new Abstract: Augmenting Transformers with linguistic structures effectively enhances the syntactic generalization performance of language models. Previous work in th

18 May 2026

PhyDetEx: Detecting and Explaining the Physical Plausibility of T2V Models

arXiv:2512.01843v2 Announce Type: replace Abstract: Driven by the growing capacity and training scale, Text-to-Video (T2V) generation models have recently achieved substantial progress in video qualit

18 May 2026

Towards Foundation Models for Relational Databases with Language Models and Graph Neural Networks

arXiv:2605.16085v1 Announce Type: cross Abstract: Relational databases store much of the world's structured information, and they are essential for driving complex predictive applications. However, de

researcharxiv-cs-ai

18 May 2026

Agentic Systems as Boosting Weak Reasoning Models

arXiv:2605.14163v1 Announce Type: new Abstract: Can a committee of weak reasoning-model calls reach the performance of much stronger models? We study verifier-backed committee search as inference-time

Safety

Do Language Models Align with Brains? Prediction Scores Are Not Enough

arXiv:2605.14025v1 Announce Type: cross Abstract: Brain-language model comparisons often interpret neural prediction scores as evidence that model representations capture brain-relevant language compu

safetyarxiv-cs-ai

PALMS: A Computational Implementation for Pavlovian Associative Learning Models' Simulation

arXiv:2602.07519v3 Announce Type: replace Abstract: In contrast to static formalisms, computational definitions describe the operational mechanisms of a model. Simulations are an essential part of the

researcharxiv-cs-lg

Probing into Camera Control of Video Models

arXiv:2605.14815v1 Announce Type: new Abstract: Video is a rich and scalable source of 3D/4D visual observations, and camera control is a key capability for video generation models to produce geometri

Energy Scaling Laws for Diffusion Models: Quantifying Compute in Image Generation

arXiv:2511.17031v2 Announce Type: replace-cross Abstract: The rapidly growing computational demands of diffusion models for image generation have raised significant concerns about energy consumption a

14 May 2026

Negation Neglect: When models fail to learn negations in training

arXiv:2605.13829v1 Announce Type: cross Abstract: We introduce Negation Neglect, where finetuning LLMs on documents that flag a claim as false makes them believe the claim is true. For example, models

14 May 2026

Multi-Narrow Transformation as a Single-Model Ensemble: Boundary Conditions, Mechanisms, and Failure Modes

arXiv:2605.11530v1 Announce Type: new Abstract: Single-model ensembles (SMEs) have attracted attention as a way to approximate some of the benefits of deep ensembles within a single network. However,

13 May 2026

Test-Time Compute for Dense Retrieval: Agentic Program Generation with Frozen Embedding Models

arXiv:2605.11374v1 Announce Type: cross Abstract: Test-time compute is widely believed to benefit only large reasoning models. We show it also helps small embedding models. Most modern embedding check

13 May 2026

ACWM-Phys: Investigating Generalized Physical Interaction in Action-Conditioned Video World Models

arXiv:2605.08567v1 Announce Type: new Abstract: Action-conditioned world models (ACWMs) have shown strong promise for video prediction and decision-making. However, existing benchmarks are largely res

CapVector: Learning Transferable Capability Vectors in Parametric Space for Vision-Language-Action Models

arXiv:2605.10903v1 Announce Type: new Abstract: This paper proposes a novel approach to address the challenge that pretrained VLA models often fail to effectively improve performance and reduce adapta

Compact SO(3) Equivariant Atomistic Foundation Models via Structural Pruning

arXiv:2605.08885v1 Announce Type: new Abstract: SO(3) equivariant graph neural networks have become the dominant paradigm for atomistic foundation models, achieving high accuracy and data efficiency b

researcharxiv-cs-lg

Tutorials

Developing a foundation model for high-resolution remote sensing data of the Netherlands

arXiv:2605.10184v1 Announce Type: cross Abstract: We develop a foundation model using 1.2m high resolution satellite images of the Netherlands. By combining a Convolutional Neural Network and a Vision

tutorialsarxiv-cs-ai

From Pixels to Concepts: Do Segmentation Models Understand What They Segment?

arXiv:2605.09591v1 Announce Type: new Abstract: Segmentation is a fundamental vision task underlying numerous downstream applications. Recent promptable segmentation models, such as Segment Anything M

Applications

HarmoWAM: Harmonizing Generalizable and Precise Manipulation via Adaptive World Action Models

arXiv:2605.10942v1 Announce Type: new Abstract: World Action Models (WAMs) have emerged as a promising paradigm for robot control by modeling physical dynamics. Current WAMs generally follow two parad

applicationsarxiv-cs-ro

LegalCiteBench: Evaluating Citation Reliability in Legal Language Models

arXiv:2605.10186v1 Announce Type: cross Abstract: Large language models (LLMs) are increasingly integrated into legal drafting and research workflows, where incorrect citations or fabricated precedent

Hardware

Model-Aware Tokenizer Transfer

arXiv:2510.21954v2 Announce Type: replace Abstract: Large Language Models (LLMs) are trained to support an increasing number of languages, yet their predefined tokenizers remain a bottleneck for adapt

hardwarearxiv-cs-cl

Scaling Vision Models Does Not Consistently Improve Localisation-Based Explanation Quality

arXiv:2605.10142v1 Announce Type: cross Abstract: Artificial intelligence models are increasingly scaled to improve predictive accuracy, yet it remains unclear whether scale improves the quality of po

Domain-level metacognitive monitoring in frontier LLMs: A 33-model atlas

arXiv:2605.06673v1 Announce Type: cross Abstract: Aggregate metacognitive quality scores mask within-model variation across MMLU benchmark domains. We administered 1,500 MMLU items (250 per domain, un

11 May 2026

Position: the Stochastic Parrot in the Coal Mine. Model Collapse is a Threat to Low-Resource Communities

arXiv:2605.04127v1 Announce Type: cross Abstract: Model collapse, the degradation in performance that arises when generative models are trained on the outputs of prior models, is an increasing concern

researcharxiv-cs-cl

7 May 2026

Replacing Parameters with Preferences: Federated Alignment of Heterogeneous Vision-Language Models

arXiv:2605.03426v1 Announce Type: new Abstract: Vision-Language Models (VLMs) have broad potential in privacy-sensitive domains such as healthcare and finance, yet strict data-sharing constraints rend