ArXiv Intelligence

Risk Under Pressure: Compute-Aware Evaluation of Adversarial Robustness in Language Models

Topic · 其他

仅有原始 MD

Quick Read

LLM failed, fallback used

Steering Where to Listen: Instruction-Based Activation Steering Redirects Temporal Attention in Large Audio-Language Models

Topic · 其他

仅有原始 MD

Quick Read

LLM failed, fallback used

详情问答

Small Experiments, Cheaper Decisions: A Case Study in Staged Promotion for Micro-Pretraining

Topic · 大模型底座

仅有原始 MD

Quick Read

LLM failed, fallback used

详情问答

Overcoming State Inertia in Full-Duplex Spoken Language Models via Activation Steering

Topic · 其他

仅有原始 MD

Quick Read

LLM failed, fallback used

详情问答

When Probing Accuracy Saturates, Fragility Resolves: A Complementary Metric for LLM Pre-Training Analysis

Topic · 其他

仅有原始 MD

Quick Read

LLM failed, fallback used

详情问答

The Dynamics of Human and AI-Generated Language: How Semantics Fluctuates across Different Timescales

Topic · 其他

仅有原始 MD

Quick Read

LLM failed, fallback used

详情问答

TileFuse: A Fused Mixed-Precision Kernel Library for Efficient Quantized LLM Inference on AMD NPUs

Topic · 机器学习框架

仅有原始 MD

Quick Read

LLM failed, fallback used

详情问答

Quantized Stochastic Primal-Dual Methods for Distributed Optimization under Relaxed Global Geometry

Topic · 其他

仅有原始 MD

Quick Read

LLM failed, fallback used

详情问答

Embodied-R1.5: Evolving Physical Intelligence via Embodied Foundation Models

Topic · 具身智能

仅有原始 MD

Quick Read

LLM failed, fallback used

详情问答

FlowBank: Query-Adaptive Agentic Workflows Optimization through Precompute-and-Reuse

Topic · Agent

仅有原始 MD

Quick Read

LLM failed, fallback used

详情问答

FreeBridge: Variational Schrödinger Bridges for Cellular Transition Dynamics

Topic · 其他

仅有原始 MD

Quick Read

LLM failed, fallback used

详情问答

RoVE: Rotary Value Embeddings Attention for Relative Position-dependent Value Pathways

Topic · 其他

仅有原始 MD

Quick Read

LLM failed, fallback used

详情问答

Federated continual learning: A comprehensive survey on lifelong and privacy-preserving learning over distributed and non-stationary data

Topic · 其他

仅有原始 MD

Quick Read

LLM failed, fallback used

详情问答

Quantifying Subliminal Behavioral Transfer Ratios in Language Model Distillation

Topic · 其他

仅有原始 MD

Quick Read

LLM failed, fallback used

详情问答

When Poison Fails After Retrieval: Revisiting Corpus Poisoning under Chunking and Reranking Pipelines

Topic · 其他

仅有原始 MD

Quick Read

LLM failed, fallback used

详情问答

OmniBioTwin: A System-of-Twinned-Systems Framework for Health Digital Twins

Topic · 机器学习框架

仅有原始 MD

Quick Read

LLM failed, fallback used

详情问答

PermDoRA -- Understanding Adapter Interference in Language Models: Limits of Parameter-Space Geometry

Topic · 其他

仅有原始 MD

Quick Read

LLM failed, fallback used

详情问答

RAIL: Rethinking Auditory Intelligence in Large Audio-Language Models with a CHC-Grounded Benchmark

Topic · 其他

仅有原始 MD

Quick Read

LLM failed, fallback used

详情问答

Physics-informed generative AI for semiconductor manufacturing: Enforcing hard physical constraints in generative models by construction

Topic · 其他

仅有原始 MD

Quick Read

LLM failed, fallback used

详情问答

SPEAR: A System for Post-Quantization Error-Adaptive Recovery Enabling Efficient Low-Bit LLM Serving

Topic · 其他

仅有原始 MD

Quick Read

LLM failed, fallback used

详情问答

2026-06-11 · 199 篇

Risk Under Pressure: Compute-Aware Evaluation of Adversarial Robustness in Language Models

Steering Where to Listen: Instruction-Based Activation Steering Redirects Temporal Attention in Large Audio-Language Models

Small Experiments, Cheaper Decisions: A Case Study in Staged Promotion for Micro-Pretraining

Overcoming State Inertia in Full-Duplex Spoken Language Models via Activation Steering

When Probing Accuracy Saturates, Fragility Resolves: A Complementary Metric for LLM Pre-Training Analysis

The Dynamics of Human and AI-Generated Language: How Semantics Fluctuates across Different Timescales

TileFuse: A Fused Mixed-Precision Kernel Library for Efficient Quantized LLM Inference on AMD NPUs

Quantized Stochastic Primal-Dual Methods for Distributed Optimization under Relaxed Global Geometry

Embodied-R1.5: Evolving Physical Intelligence via Embodied Foundation Models

FlowBank: Query-Adaptive Agentic Workflows Optimization through Precompute-and-Reuse

FreeBridge: Variational Schrödinger Bridges for Cellular Transition Dynamics

RoVE: Rotary Value Embeddings Attention for Relative Position-dependent Value Pathways

Federated continual learning: A comprehensive survey on lifelong and privacy-preserving learning over distributed and non-stationary data

Quantifying Subliminal Behavioral Transfer Ratios in Language Model Distillation

When Poison Fails After Retrieval: Revisiting Corpus Poisoning under Chunking and Reranking Pipelines

OmniBioTwin: A System-of-Twinned-Systems Framework for Health Digital Twins

PermDoRA -- Understanding Adapter Interference in Language Models: Limits of Parameter-Space Geometry

RAIL: Rethinking Auditory Intelligence in Large Audio-Language Models with a CHC-Grounded Benchmark

Physics-informed generative AI for semiconductor manufacturing: Enforcing hard physical constraints in generative models by construction

SPEAR: A System for Post-Quantization Error-Adaptive Recovery Enabling Efficient Low-Bit LLM Serving