Search: “benchmark”

Right Knowledge, Wrong Answer: Test-Time Steering for Temporal Fact Conflicts in Open-Weight Language Models

arXiv:2606.20959v1 Announce Type: new Abstract: Large language models can store both outdated facts and newer superseding facts in their parameters, but standard prompting may still elicit the outdate

Content type

All Blog X Post Paper YouTube Reddit GitHub

Clear filters

Local Ai

RIZZ: Routing Interactions to Near Zero-Interference Zones for Continual Adaptation of Black-Box Agents

DGX agent

arXiv:2606.20638v1 Announce Type: cross Abstract: Large language models are increasingly deployed as long-lived agents that must adapt across users, tasks, domains, modalities, and feedback regimes wi

local-aiarxiv-cs-lg

23 Jun 2026

RLM-Cascade: Response-Level Speculative Decoding for Cost-Efficient LLM API Serving

arXiv:2606.22840v1 Announce Type: new Abstract: We present RLM-Cascade, a proxy-layer system that applies speculative decoding at the response level to reduce LLM API costs without requiring model arc

RS-Gen: A Multi-Stage Agentic Framework for Reasoning and Search-Augmented Image Generation

arXiv:2606.23221v1 Announce Type: new Abstract: Recent years have witnessed remarkable progress in image generation and editing, particularly regarding instruction following and visual fidelity. Howev

RT-DocLayout: Real-Time End-to-End Document Layout Analysis with Reading Order in the Wild

arXiv:2606.23344v1 Announce Type: new Abstract: Accurate document layout analysis remains a critical bottleneck for document parsing systems, due to the intricate coupling among heterogeneous document

Tutorials

Scalable Training of Spatially Grounded 2D Vision-Language Models for Radiology

arXiv:2606.20477v2 Announce Type: replace Abstract: We study how to train visually grounded vision-language models (VLMs) for radiology without manual spatial annotations. We introduce RefRad2D, a lar

tutorialsarxiv-cs-cv

Local Ai

Scaling Diverse Language Generation for 3D Visual Grounding

arXiv:2606.20946v1 Announce Type: cross Abstract: Developing robust models for 3D visual grounding (3DVG), the localization of entities in a 3D scene described in natural language, is important for en

local-aiarxiv-cs-cv

Scaling Self-Play for End-to-End Driving

arXiv:2606.19641v2 Announce Type: replace-cross Abstract: End-to-end autonomous driving models are typically trained on offline human-demonstration datasets that provide limited state coverage and oft

safetyarxiv-cs-cv

scLLM-DSC: LLM-Knowledge Enhanced Cross-Modal Deep Structural Clustering for Single-Cell RNA Sequencing

arXiv:2606.13007v2 Announce Type: replace Abstract: Clustering is fundamental to scRNA-seq analysis, serving as a cornerstone for identifying cell populations and resolving tissue heterogeneity. Howev

safetyarxiv-cs-lg

Seeing Before Reasoning: Decoupling Perception and Reasoning for Shortcut-Resilient Multimodal On-Policy Self-Distillation

arXiv:2606.19120v2 Announce Type: replace-cross Abstract: On-policy self-distillation (OPSD) trains a model on its own rollouts and uses a frozen copy to provide dense token-level targets conditioned

safetyarxiv-cs-cv

SeFi-Image: A Text-to-Image Foundation Model with Semantic-First Diffusion

arXiv:2606.22568v1 Announce Type: new Abstract: Training image generation foundation models consumes substantial resources. Previous methods have attempted to leverage semantic guidance to accelerate

Select-to-Act: Hierarchical Reinforcement Learning via Adaptive Language Guidance

arXiv:2606.22350v1 Announce Type: new Abstract: Reinforcement Learning (RL) has been widely applied to sequential decision-making, yet it often suffers from poor sample efficiency due to costly intera

Self-Evolution for Multi-Turn Tool-Calling Agents via Divergence-Point Preference Learning

arXiv:2606.23112v1 Announce Type: new Abstract: Multi-turn tool-using agents must coordinate long-horizon tool sequences while tracking dialogue state and policy constraints. Existing approaches often

Self-Evolving Cognitive Framework via Causal World Modeling for Embodied Scientific Intelligence

arXiv:2606.22449v1 Announce Type: cross Abstract: Current embodied world models are primarily optimized for predictive objectives, limiting their ability to generalize under distribution shifts and re

researcharxiv-cs-ro

Sequential Minimal Optimization Algorithm for One-Class Support Vector Machines With Privileged Information

arXiv:2606.22210v1 Announce Type: new Abstract: One of the powerful techniques in data modeling is accounting for features that are available at the training stage, but are not available when the trai

Set-based v.s. Distribution-based Representations of Epistemic Uncertainty: A Comparative Study

arXiv:2602.22747v2 Announce Type: replace Abstract: Epistemic uncertainty in neural networks is commonly modeled using two second-order paradigms: distribution-based representations, which rely on pos

Agents

Sim2O: Efficient Offline-to-Online MARL via Joint Action Composition

arXiv:2606.21085v1 Announce Type: new Abstract: Offline-to-online adaptation serves as a pivotal paradigm for mitigating the prohibitive cost of online exploration by bootstrapping reinforcement learn

agentsarxiv-cs-lg

SimAC: A Simple Anti-Customization Method for Protecting Face Privacy against Text-to-Image Synthesis of Diffusion Models

arXiv:2312.07865v4 Announce Type: replace Abstract: Despite the success of diffusion-based customization methods on visual content creation, increasing concerns have been raised about such techniques

SOHET: Sequence Of Heterogeneous Events Transformer with Self-Supervised Pre-Training

arXiv:2606.21356v1 Announce Type: new Abstract: Many machine learning applications rely on heterogeneous event streams to make predictions, either causally as events arrive or bidirectionally over com

SparseWorld: Enhancing End-to-End Autonomous Driving via World Models with Sparse Scene Representation

arXiv:2605.24354v2 Announce Type: replace Abstract: Recently, world models have made significant progress in enhancing end-to-end driving systems through both future situation forecasting and improved

Local Ai

SPOT-E: Test-Time Entropy Shaping with Visual Spotlights for Frozen VLMs

arXiv:2606.20244v2 Announce Type: replace Abstract: Vision-language models (VLMs) often underperform on evidence intensive tasks because decisive visual evidence are small, localized, and easy to over

local-aiarxiv-cs-cv

Applications

Statistical Matching via Schrodinger Bridge beyond Conditional Independence

arXiv:2606.22770v1 Announce Type: new Abstract: Statistical matching combines partially overlapping datasets that share covariates X but observe the target Y and auxiliary variables Z separately. Clas

applicationsarxiv-cs-lg

Streaming Dense Voxel Representations for 3D Occupancy Prediction

arXiv:2503.22087v3 Announce Type: replace Abstract: In this paper, we explore dense voxel streaming for accurate and efficient 3D occupancy prediction. While dense voxel representations offer fine-gra

StructSAM: Structure- and Spectrum-Preserving Token Merging for Segment Anything Models

arXiv:2603.07307v2 Announce Type: replace Abstract: Recent token merging techniques for Vision Transformers (ViTs) provide substantial speedups by reducing the number of tokens processed by self-atten

Structured Hyperedge Adaptation for Parameter-Efficient Fine-Tuning of Vision Transformers

arXiv:2606.22383v1 Announce Type: new Abstract: Parameter-efficient fine-tuning (PEFT) has become a practical solution for adapting large pretrained vision transformers (ViTs) to downstream tasks whil

Subject-Level Unknown-Identity Identification from Leap Motion Controller 2 Hand Landmarks

arXiv:2606.22986v1 Announce Type: new Abstract: This work studies subject recognition from Leap Motion Controller 2 (LMC2) hand landmark data under a subject-level unknown-identity identification prot

Sublinearly Structured Deep Neural Networks Achieve Feature Learning Consistency for Compositional Functions

arXiv:2606.23477v1 Announce Type: cross Abstract: Over the past decade, deep neural networks (DNNs) have achieved remarkable success on complex machine-learning tasks, yet the theoretical foundations

researcharxiv-cs-lg

Surprise-Guided MergeSort: Budget-Efficient Human-in-the-Loop Ranking via Adaptive Comparison Scheduling

arXiv:2606.15623v2 Announce Type: replace Abstract: Pairwise comparison is the gold standard for subjective ranking tasks; however, exhaustive annotation requires a massive number of human comparisons

safetyarxiv-cs-lg

Synergistic Dual-Branch Adaptation for Multi-modal Generalized Category Discovery

arXiv:2606.21446v1 Announce Type: new Abstract: Generalized Category Discovery (GCD) aims to classify old categories and discover new ones from unlabeled data. Recent multi-modal approaches introduce

safetyarxiv-cs-cv

T-MOR: Learning Motion-Aware Skeleton Representations for Human Action Recognition

arXiv:2606.21607v1 Announce Type: new Abstract: Vision-language models such as CLIP have recently achieved strong performance on a wide range of visual understanding tasks. However, most existing mode

T-VSS: Test-Time Visual Subspace Steering for Adversarial Robustness of Vision-Language Models

arXiv:2606.23132v1 Announce Type: new Abstract: Vision-language models (VLMs) achieve strong zero-shot recognition, but they remain highly vulnerable to adversarial perturbations. Recent test-time ada

Applications

TaLK: Text-attributed Graph Dataset Distillation via Coupling Language Model with Graph-Aware Kernel

arXiv:2606.22975v1 Announce Type: new Abstract: Text-attributed graphs (TAGs) are widely used in many real-world domains, and learning on TAGs requires jointly modeling text semantics and graph struct

applicationsarxiv-cs-lg

Tapered Language Models

arXiv:2606.23670v1 Announce Type: new Abstract: Modern language models, including transformer, recurrent, and memory-based variants, share a common chassis: a stack of identical layers in which parame

Task-Differentiated Atomic Skill Expansion and Routing for Continual Learning Across Highly Heterogeneous Tasks

arXiv:2606.21307v1 Announce Type: new Abstract: Continual learning (CL) is commonly studied under the assumption that sequential tasks are semantically related or structurally similar. However, in hig

Technical Report for the ICRA 2026 GOOSE 2D Fine-Grained Semantic Segmentation Challenge: Pretraining-Diverse Ensemble of Foundation Vision Encoders for Robust Outdoor Scene Understanding

arXiv:2606.23113v1 Announce Type: new Abstract: This report presents our solution for the ICRA 2026 GOOSE 2D Fine-Grained Semantic Segmentation Challenge, which requires parsing unstructured outdoor s

Temper-Then-Tilt: Principled Unlearning for Generative Models through Tempering and Classifier Guidance

arXiv:2602.10217v2 Announce Type: replace Abstract: We study machine unlearning in large generative models by framing the task as density ratio estimation to a target distribution rather than supervis

Temporal Causal Prior-Data Fitted Networks for Panel Data with Learned Reliability Signals

arXiv:2606.20889v1 Announce Type: new Abstract: Estimating causal effects in industrial time series requires handling temporal dynamics, time-varying treatments, and unobserved confounders. Existing c

Temporal-Spectral Alignment with Frequency Adaptation for Source-Free Time-Series Adaptation

arXiv:2606.23120v1 Announce Type: new Abstract: The goal of source-free domain adaptation (SFDA) for time-series data is to transfer knowledge from a pre-trained source model to an unlabeled target do

Temporally Aware Densification for Dynamic 3D Gaussian Splatting

arXiv:2606.23212v1 Announce Type: new Abstract: Despite modeling temporal motion, dynamic 3D Gaussian Splatting (3DGS) methods still inherit a static densification strategy that is ill-suited for dyna

TF-SNO: Time-Frequency Gated Spectral Neural Operators for Learning Non-Stationary Partial Differential Equations

arXiv:2606.21189v1 Announce Type: new Abstract: Non-stationary partial differential equations (PDEs) arise throughout scientific computing, where the dominant frequency content and energy distribution

researcharxiv-cs-lg

Thalia: A Global, Multi-Modal Dataset for Volcanic Activity Monitoring

arXiv:2505.17782v4 Announce Type: replace Abstract: Monitoring volcanic activity is of paramount importance to safeguarding lives, infrastructure, and ecosystems. However, only a small fraction of kno

The Alignment Problem in Constrained Code Generation

arXiv:2606.21619v1 Announce Type: cross Abstract: Large Language Models (LLMs) have demonstrated strong capabilities in code generation, but their outputs frequently contain syntax or type errors that

safetyarxiv-cs-lg

The MAMA-MIA Challenge: Advancing Generalizability and Fairness in Breast MRI Tumor Segmentation and Treatment Response Prediction

arXiv:2603.01250v3 Announce Type: replace Abstract: Breast cancer is the most frequently diagnosed malignancy among women worldwide and a leading cause of cancer-related mortality. Dynamic contrast-en

The Power of Light: Improving Synthetic-to-Real Domain Adaptation through Physically-Based Indirect Illumination

arXiv:2606.22574v1 Announce Type: new Abstract: While synthetic data generation resolves the manual labeling bottleneck in computer vision, minimizing the syn-to-real domain gap requires optimizing re

The Two-Hump Problem: Bridging the Difficulty Gap in Mathematical Reinforcement Learning

arXiv:2606.21611v1 Announce Type: new Abstract: Mathematical search problems present a unique challenge for Reinforcement Learning (RL) due to vast search spaces and sparse rewards. In previous works,

The Unreasonable Effectiveness of VLMs for Zero-shot Procedural Mistake Detection

arXiv:2606.21579v1 Announce Type: new Abstract: Procedural mistake detection is important for quality control and user assistance across many disciplines. Recent work in this field has achieved signif

TooBad: Backdoor Diffusion Models with Ultra-Low Poison Rate and Imperceptible Trigger

arXiv:2606.23362v1 Announce Type: cross Abstract: Diffusion models (DMs), despite their impressive capabilities across a wide range of generative tasks, have been shown to be vulnerable to backdoor at

Topological summaries of fingerprint ridge patterns carry identity information

arXiv:2606.22029v1 Announce Type: new Abstract: Fingerprints are the most widely deployed biometric. Verifying whether two impressions come from the same finger typically relies on minutiae, small lan