Search: “reasoning”

arXiv:2605.19738v1 Announce Type: cross Abstract: Graph Anomaly Detection (GAD) aims to identify atypical graph entities, such as nodes, edges, or substructures, that deviate significantly from the ma

applicationsarxiv-cs-ai

Content type

All Blog X Post Paper YouTube Reddit GitHub

Clear filters

Safety

The Accessibility Capability Boundary: Operational Limits and Expansion Potential of AI-Generated Browser-Native Accessibility Systems

DGX agent

arXiv:2605.19638v1 Announce Type: cross Abstract: As large language models (LLMs) demonstrate increasing competence in synthesizing functional user interfaces, a fundamental question emerges in access

safetyarxiv-cs-ai

20 May 2026

ThoughtTrace: Understanding User Thoughts in Real-World LLM Interactions

arXiv:2605.20087v1 Announce Type: cross Abstract: Conversational AI has now reached billions of users, yet existing datasets capture only what people say, not what they think. We introduce ThoughtTrac

Towards Data-Efficient Video Pre-training with Frozen Image Foundation Models

arXiv:2605.19137v1 Announce Type: new Abstract: Video foundation models achieve strong performance across many video understanding tasks, but typically require large-scale pre-training on massive vide

TravExplorer: Cross-Floor Embodied Exploration via Traversability-Aware 3-D Planning

model-releasesarxiv-cs-ro

arXiv:2605.19958v1 Announce Type: new Abstract: Zero-shot Object Navigation (ZSON) has shown promise for open-vocabulary target search in unseen environments, yet most existing systems remain tied to

Unlocking the Potential of Continual Model Merging: An ODE Perspective

arXiv:2605.19409v1 Announce Type: cross Abstract: Continual Model Merging (CMM) enables rapid customization of foundation models across sequentially arriving tasks, offering a scalable alternative to

Where Not to Learn: Prior-Aligned Training with Subset-based Attribution Constraints for Reliable Decision-Making

arXiv:2602.07008v2 Announce Type: replace Abstract: Reliable models should not only predict correctly, but also justify decisions with acceptable evidence. Yet conventional supervised learning typical

safetyarxiv-cs-cv

Applications

A Conflict-aware Evidential Framework for Reliable Sleep Stage Classification

arXiv:2605.17021v1 Announce Type: new Abstract: Multi-view learning has been widely applied for sleep stage classification using multi-modal data. However, existing methods typically assume that diffe

applicationsarxiv-cs-ai

A-ProS: Towards Reliable Autonomous Programming Through Multi-Model Feedback

arXiv:2605.18073v1 Announce Type: cross Abstract: Large Language Models (LLMs) demonstrate strong potential for automated code generation, yet their ability to iteratively refine solutions using execu

AMATA: Adaptive Multi-Agent Trajectory Alignment for Knowledge-Intensive Question Answering

arXiv:2605.17352v1 Announce Type: new Abstract: Despite substantial advances in large language models (LLMs), generating factually consistent responses for knowledge-intensive question answering remai

safetyarxiv-cs-cl

An Efficient Streaming Video Understanding Framework with Agentic Control

arXiv:2605.17921v1 Announce Type: new Abstract: Streaming video requires handling dynamic information density under strict latency budgets. Yet, existing methods typically employ static strategies, su

safetyarxiv-cs-cv

Artificial Intolerance: Stigmatizing Language in Clinical Documentation Skews Large Language Model Decision-Making

arXiv:2605.17228v1 Announce Type: new Abstract: Large Language Models (LLMs) are increasingly deployed in high-stakes domains such as clinical decision support and medical documentation. However, the

safetyarxiv-cs-cl

ArtMesh: Part-Aware Articulated Mesh Fields with Motion-Consistent Dynamics

arXiv:2605.16582v1 Announce Type: new Abstract: We present ArtMesh, a mesh-native method for reconstructing articulated objects explicitly as connected triangle meshes with per-part rigid motion from

Asking Back: Interaction-Layer Antidistillation Watermarks

arXiv:2605.16462v1 Announce Type: cross Abstract: Detecting unauthorized knowledge distillation from a deployed LLM API is hard because the defender controls neither the attacker's training pipeline n

Tutorials

Automated Knowledge Component Generation for Interpretable Knowledge Tracing in Coding Problems

arXiv:2502.18632v4 Announce Type: replace Abstract: Knowledge components (KCs) mapped to problems help model student learning, tracking their mastery levels on fine-grained skills thereby facilitating

tutorialsarxiv-cs-ai

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

Beacon: Single-Turn Diagnosis and Mitigation of Latent Sycophancy in Large Language Models

arXiv:2510.16727v2 Announce Type: replace-cross Abstract: Large language models internalize a structural trade-off between truthfulness and obsequious flattery, emerging from reward optimization that

BELIEF: Structured Evidence Modeling and Uncertainty-Aware Fusion for Biomedical Question Answering

arXiv:2605.17435v1 Announce Type: new Abstract: Biomedical question answering often requires decisions from retrieved literature whose relevance, quality, and support for candidate answers are uneven.

researcharxiv-cs-cl

Local Ai

Beyond Inference-Time Search: Reinforcement Learning Synthesizes Reusable Solvers

arXiv:2605.18374v1 Announce Type: cross Abstract: Large language models (LLMs) typically approach combinatorial optimization as an inference-time procedure, solving each instance separately through sa

local-aiarxiv-cs-ai

CAM-VFD: Cross-Attention Multimodal Video Forgery Detection

arXiv:2605.17133v1 Announce Type: cross Abstract: The rapid advancement of Deepfake technologies and video manipulation tools poses a critical challenge to multimedia forensics, judicial evidence inte

Can LLMs Think Like Consumers? Benchmarking Crowd-Level Reaction Reconstruction with ConsumerSimBench

arXiv:2605.17079v1 Announce Type: cross Abstract: LLMs are increasingly used as ``digital consumers'' to simulate public opinion, pre-test marketing decisions, and anticipate audience response. Howeve

CodeScaler: Scaling Code LLM Training and Test-Time Inference via Reward Models

arXiv:2602.17684v2 Announce Type: replace-cross Abstract: Reinforcement Learning from Verifiable Rewards (RLVR) has driven recent progress in code large language models by leveraging execution-based f

Continual Learning for VLMs: A Survey and Taxonomy Beyond Forgetting

arXiv:2508.04227v2 Announce Type: replace Abstract: Vision-language models (VLMs) and the recent surge of Multimodal Large Language Models (MLLMs) have revolutionized artificial intelligence with unpr

Convex Dataset Valuation for Post-Training

arXiv:2605.16704v1 Announce Type: new Abstract: Improving LLM performance on downstream tasks sometimes requires leveraging auxiliary datasets during post-training. In practice, however, developers fa

safetyarxiv-cs-lg

Counterparty Modeling is Not Strategy: The Limits of LLM Negotiators

arXiv:2605.16575v1 Announce Type: new Abstract: Negotiation requires more than inferring what the other side wants: it requires using that information to make advantageous offers and counteroffers ove

CyberCorrect: A Cybernetic Framework for Closed-Loop Self-Correction in Large Language Models

arXiv:2605.17305v1 Announce Type: new Abstract: Large language model (LLM) self-correction -- the ability to detect and fix errors in generated outputs -- remains largely ad hoc, relying on generic pr

Deep sequence models tend to memorize geometrically; it is unclear why

arXiv:2510.26745v3 Announce Type: replace-cross Abstract: Deep sequence models are said to store atomic facts predominantly in the form of associative memory: a brute-force lookup of co-occurring enti

DevBench: A Realistic, Developer-Informed Benchmark for Code Generation Models

arXiv:2601.11895v3 Announce Type: replace-cross Abstract: DevBench is a telemetry-driven benchmark designed to evaluate Large Language Models (LLMs) on realistic code completion tasks. It includes 1,8

Dynamic robotic cloth folding with efficient Koopman operator-based model predictive control

arXiv:2605.18373v1 Announce Type: cross Abstract: Robotic cloth folding is a challenging task, particularly when considering dynamic folding tasks, which aim at folding cloth by fast motions that leve

researcharxiv-cs-lg

Echoes in Filter Bubble: Diagnosing and Curing Popularity Bias in Generative Recommenders

arXiv:2605.16825v1 Announce Type: cross Abstract: Recently, Generative Recommenders (GRs), characterized by a unified end-to-end framework, have exhibited astonishing potential in transforming the rec

Agents

EnvFactory: Scaling Tool-Use Agents via Executable Environments Synthesis and Robust RL

arXiv:2605.18703v1 Announce Type: new Abstract: Equipping LLMs with tool-use capabilities via Agentic Reinforcement Learning (Agentic RL) is bottlenecked by two challenges: the lack of scalable, robus

agentsarxiv-cs-cl

EPIC-Bench: A Perception-Centric Benchmark for Fine-Grained Embodied Visual Grounding in Vision-Language Models

arXiv:2605.17070v1 Announce Type: new Abstract: While large vision-language models (VLMs) are increasingly adopted as the perceptual backbone for embodied agents, existing benchmarks often rely on que

Episodic-Semantic Memory Architecture for Long-Horizon Scientific Agents

arXiv:2605.17625v1 Announce Type: new Abstract: As Large Language Models (LLMs) evolve into persistent scientific collaborators, context window saturation has emerged as a critical bottleneck. Scienti

EVA01: Unified Native 3D Understanding and Generation via Mixture-of-Transformers

arXiv:2605.16745v1 Announce Type: new Abstract: This paper addresses the challenge of integrating 3D meshes as a native modality within Multimodal Large Language Models (MLLMs). Diffusion-based large

Evaluating Deep Research Agents on Expert Consulting Work: A Benchmark with Verifiers, Rubrics, and Cognitive Traps

arXiv:2605.17554v1 Announce Type: new Abstract: Frontier deep research agents (DRAs) plan a research task, synthesize across documents, and return a structured deliverable on demand. They are being de

EveryQuery: Zero-Shot Clinical Prediction via Task-Conditioned Pretraining over Electronic Health Records

arXiv:2603.07900v2 Announce Type: replace Abstract: Foundation models pretrained on electronic health records (EHR) have demonstrated zero-shot clinical prediction capabilities by generating synthetic

extit{Don't Guess, Just Ask}: Resolving Ambiguity in Referring Segmentation via Multi-turn Clarification

arXiv:2605.17531v1 Announce Type: new Abstract: Referring segmentation aims to segment the target objects in images or videos based on the textual query. Despite remarkable progress over the past year

extsc{PrivScope}: Task-scoped Disclosure Control for Hybrid Agentic Systems

arXiv:2605.16630v1 Announce Type: cross Abstract: Hybrid local--cloud agents enrich user requests with context from persistent working state before delegating capability-intensive subtasks to a cloud

Factorized Latent Dynamics for Video JEPA: An Empirical Study of Auxiliary Objectives

arXiv:2605.17165v1 Announce Type: new Abstract: Joint-Embedding Predictive Architectures (JEPA) are a promising framework for self-supervised video representation learning, yet the behavior of auxilia

Tutorials

Federated Nested Learning: Collaborative Training of Self-Referential Memories for Test-Time Adaptation

arXiv:2605.16350v1 Announce Type: cross Abstract: We rethink Federated Learning (FL) from a nested learning perspective, framing the core challenge as how to collaboratively learn optimization rules,

tutorialsarxiv-cs-ai

FIM-LoRA: Task-Informative Rank Allocation for LoRA via Calibration-Time Gradient-Variance Estimation

model-releasesarxiv-cs-cl

arXiv:2605.16800v1 Announce Type: cross Abstract: Low-rank adaptation (LoRA) assigns a uniform rank to every adapted weight matrix - a practical convenience that ignores a fundamental reality: differe

First Experimental Demonstration of Natural Hovering Extremum Seeking: A New Paradigm in Flapping Flight Physics

arXiv:2508.20836v3 Announce Type: replace Abstract: In this letter, we report the first experimental demonstration of the recently emerged new paradigm in hovering and flapping flight physics called (

researcharxiv-cs-ro

Forecasting Downstream Performance of LLMs With Proxy Metrics

arXiv:2605.18607v1 Announce Type: new Abstract: Progress in language model development is often driven by comparative decisions: which architecture to adopt, which pretraining corpus to use, or which

researcharxiv-cs-cl

FormuLLA: A Large Language Model Approach to Generating Novel 3D Printable Formulations

arXiv:2601.02071v3 Announce Type: replace Abstract: Pharmaceutical three-dimensional (3D) printing is an advanced fabrication technology with the potential to enable truly personalised dosage forms. R

Full Attention Strikes Back: Transferring Full Attention into Sparse within Hundred Training Steps

arXiv:2605.16928v1 Announce Type: cross Abstract: Long-context inference in large language models is bottlenecked by the quadratic cost of full attention. Existing efficient alternatives often rely ei

GaussianDWM: 3D Gaussian Driving World Model for Unified Scene Understanding and Multi-Modal Generation

arXiv:2512.23180v3 Announce Type: replace Abstract: Driving World Models (DWMs) have been developing rapidly with the advances of generative models. However, existing DWMs lack 3D scene understanding

safetyarxiv-cs-cv

GaussianZoom: Progressive Zoom-in Generative 3D Gaussian Splatting with Geometric and Semantic Guidance

arXiv:2605.18252v1 Announce Type: new Abstract: We introduce GaussianZoom, a generative zoom-in 3D reconstruction system with an iterative progressive framework that combines geometry-consistent scene

General Preference Reinforcement Learning

model-releasesarxiv-cs-cl

arXiv:2605.18721v1 Announce Type: cross Abstract: Post-training has split large language model (LLM) alignment into two largely disconnected tracks. Online reinforcement learning (RL) with verifiable