Search: “llm”

arXiv:2604.26837v1 Announce Type: new Abstract: Long-context LLM serving is bottlenecked by the cost of attending over ever-growing KV caches. Dynamic sparse attention promises relief by accessing onl

local-aiarxiv-cs-lg

30 Apr 2026

Content type

All Blog X Post Paper YouTube Reddit GitHub

Clear filters

Local Ai

MobileLLM-Flash: Latency-Guided On-Device LLM Design for Industry Scale Deployment

DGX agent

arXiv:2603.15954v2 Announce Type: replace Abstract: Real-time AI experiences call for on-device large language models (OD-LLMs) optimized for efficient deployment on resource-constrained hardware. The

local-aiarxiv-cs-lg

29 Apr 2026

Pimp My LLM: Leveraging Variability Modeling to Tune Inference Hyperparameters

arXiv:2602.17697v2 Announce Type: replace Abstract: Large Language Models (LLMs) are being increasingly used across a wide range of tasks. However, their substantial computational demands raise concer

researcharxiv-cs-lg

29 Apr 2026

A Survey on Split Learning for LLM Fine-Tuning: Models, Systems, and Privacy Optimizations

arXiv:2604.24468v1 Announce Type: cross Abstract: Fine-tuning unlocks large language models (LLMs) for specialized applications, but its high computational cost often puts it out of reach for resource

Benchmarking Emergent Coordination in Large-Scale LLM Populations: An Evaluation Framework on the MoltBook Archive

arXiv:2603.03555v2 Announce Type: replace-cross Abstract: As multi-agent Large Language Model (LLM) systems scale, evaluating their emergent coordination dynamics becomes increasingly critical. Howeve

DPEPO: Diverse Parallel Exploration Policy Optimization for LLM-based Agents

arXiv:2604.24320v1 Announce Type: new Abstract: Large language model (LLM) agents that follow the sequential 'reason-then-act' paradigm have achieved superior performance in many complex tasks.However

safetyarxiv-cs-cl

From Similarity to Structure: Training-free LLM Context Compression with Hybrid Graph Priors

arXiv:2604.23277v1 Announce Type: cross Abstract: Long-context large language models remain computationally expensive to run and often fail to reliably process very long inputs, which makes context co

KLong: Training LLM Agent for Extremely Long-horizon Tasks

arXiv:2602.17547v3 Announce Type: replace Abstract: This paper introduces KLong, an open-source LLM agent trained to solve extremely long-horizon tasks. The principle is to first cold-start the model

LEGO: An LLM Skill-Based Front-End Design Generation Platform

arXiv:2604.23355v1 Announce Type: new Abstract: Existing LLM-based EDA agents are often isolated task-specific systems. This leads to repeated engineering effort and limited reuse of successful design

One Size Fits None: Heuristic Collapse in LLM Investment Advice

arXiv:2604.23837v1 Announce Type: new Abstract: Large language models are increasingly deployed as advisors in high-stakes domains -- answering medical questions, interpreting legal documents, recomme

applicationsarxiv-cs-cl

Personality Shapes Gender Bias in Persona-Conditioned LLM Narratives Across English and Hindi: An Empirical Investigation

arXiv:2604.23600v1 Announce Type: new Abstract: Large Language Models (LLMs) are increasingly deployed in persona-driven applications such as education, customer service, and social platforms, where m

safetyarxiv-cs-cl

RouteGuard: Internal-Signal Detection of Skill Poisoning in LLM Agents

arXiv:2604.22888v1 Announce Type: cross Abstract: Agent skills introduce a new and more severe form of indirect injection for LLM agents: unlike traditional indirect prompt injection, attackers can hi

safetyarxiv-cs-ai

SFT-then-RL Outperforms Mixed-Policy Methods for LLM Reasoning

arXiv:2604.23747v1 Announce Type: cross Abstract: Recent mixed-policy optimization methods for LLM reasoning that interleave or blend supervised and reinforcement learning signals report improvements

Small Language Model Helps Resolve Semantic Ambiguity of LLM Prompt

arXiv:2604.23263v1 Announce Type: cross Abstract: Large language models (LLMs) are increasingly utilized in various complex reasoning tasks due to their excellent instruction following capability. How

The Override Gap: A Magnitude Account of Knowledge Conflict Failure in Hypernetwork-Based Instant LLM Adaptation

arXiv:2604.23750v1 Announce Type: cross Abstract: Hypernetwork-based methods such as Doc-to-LoRA internalize a document into an LLM's weights in a single forward pass, but they fail systematically on

Towards High-Fidelity CAD Generation via LLM-Driven Program Generation and Text-Based B-Rep Primitive Grounding

arXiv:2603.11831v2 Announce Type: replace Abstract: The field of Computer-Aided Design (CAD) generation has made significant progress in recent years. Existing methods typically fall into two separate

applicationsarxiv-cs-cv

What Prompts Don't Say: Understanding and Managing Underspecification in LLM Prompts

arXiv:2505.13360v3 Announce Type: replace Abstract: Prompt underspecification is a common challenge when interacting with LLMs. In this paper, we present an in-depth analysis of this problem, showing

Your Reviews Replicate You: LLM-Based Agents as Customer Digital Twins for Conjoint Analysis

arXiv:2604.22756v1 Announce Type: cross Abstract: Conjoint analysis is a cornerstone of market research for estimating consumer preferences; however, traditional methods face persistent challenges reg

applicationsarxiv-cs-ai

Chain-of-Memory: Lightweight Memory Construction with Dynamic Evolution for LLM Agents

arXiv:2601.14287v2 Announce Type: replace Abstract: External memory systems are pivotal for enabling Large Language Model (LLM) agents to maintain persistent knowledge and perform long-horizon decisio

researcharxiv-cs-lg

27 Apr 2026

Tracing the complexity profiles of different linguistic phenomena through the intrinsic dimension of LLM representations

arXiv:2601.03779v2 Announce Type: replace Abstract: We explore intrinsic dimension (ID) of LLM representations as a marker of linguistic complexity. Specifically, we test whether ID differences across

27 Apr 2026

When Does LLM Self-Correction Help? A Control-Theoretic Markov Diagnostic and Verify-First Intervention

arXiv:2604.22273v1 Announce Type: new Abstract: Iterative self-correction is widely used in agentic LLM systems, but when repeated refinement helps versus hurts remains unclear. We frame self-correcti

27 Apr 2026

A Replicable Robotics Awareness Method Using LLM-Enabled Robotics Interaction: Evidence from a Corporate Challenge

arXiv:2604.21377v1 Announce Type: new Abstract: Large language models are increasingly being explored as interfaces between humans and robotic systems, yet there remains limited evidence on how such t

researcharxiv-cs-ro

CI-Work: Benchmarking Contextual Integrity in Enterprise LLM Agents

model-releasesarxiv-cs-cl

arXiv:2604.21308v1 Announce Type: cross Abstract: Enterprise LLM agents can dramatically improve workplace productivity, but their core capability, retrieving and using internal context to act on a us

Clinically Interpretable Sepsis Early Warning via LLM-Guided Simulation of Temporal Physiological Dynamics

arXiv:2604.20924v1 Announce Type: new Abstract: Timely and interpretable early warning of sepsis remains a major clinical challenge due to the complex temporal dynamics of physiological deterioration.

agentsarxiv-cs-lg

Co-Evolving LLM Decision and Skill Bank Agents for Long-Horizon Tasks

arXiv:2604.20987v1 Announce Type: new Abstract: Long horizon interactive environments are a testbed for evaluating agents skill usage abilities. These environments demand multi step reasoning, the cha

agentsarxiv-cs-ai

DryRUN: On the Role of Public Tests in LLM-Driven Code Generation

arXiv:2604.21598v1 Announce Type: cross Abstract: Multi-agent frameworks are widely used in autonomous code generation and have applications in complex algorithmic problem-solving. Recent work has add

agentsarxiv-cs-ai

InfiniPipe: Elastic Pipeline Parallelism for Efficient Variable-Length Long-Context LLM Training

arXiv:2509.21275v3 Announce Type: replace-cross Abstract: Long context training is crucial for LLM's context extension. Existing schemes, such as sequence parallelism, incur substantial communication

applicationsarxiv-cs-ai

Job Skill Extraction via LLM-Centric Multi-Module Framework

arXiv:2604.21525v1 Announce Type: new Abstract: Span-level skill extraction from job advertisements underpins candidate-job matching and labor-market analytics, yet generative large language models (L

Measuring Opinion Bias and Sycophancy via LLM-based Coercion

model-releasesarxiv-cs-cl

arXiv:2604.21564v1 Announce Type: new Abstract: Large language models increasingly shape the information people consume: they are embedded in search, consulted for professional advice, deployed as age

Omission Constraints Decay While Commission Constraints Persist in Long-Context LLM Agents

arXiv:2604.20911v1 Announce Type: cross Abstract: LLM agents deployed in production operate under operator-defined behavioral policies (system-prompt instructions such as prohibitions on credential di

Local Ai

SparKV: Overhead-Aware KV Cache Loading for Efficient On-Device LLM Inference

arXiv:2604.21231v1 Announce Type: cross Abstract: Efficient inference for on-device Large Language Models (LLMs) remains challenging due to limited hardware resources and the high cost of the prefill

local-aiarxiv-cs-ai

Zero-Shot Detection of LLM-Generated Text via Implicit Reward Model

arXiv:2604.21223v1 Announce Type: cross Abstract: Large language models (LLMs) have demonstrated remarkable capabilities across various tasks. However, their ability to generate human-like text has ra

CAST: Achieving Stable LLM-based Text Analysis for Data Analytics

arXiv:2602.15861v2 Announce Type: replace-cross Abstract: Text analysis of tabular data relies on two core operations: summarization for corpus-level theme extraction and tagging for row-level labelin

safetyarxiv-cs-ai

ESGLens: An LLM-Based RAG Framework for Interactive ESG Report Analysis and Score Prediction

arXiv:2604.19779v1 Announce Type: new Abstract: Environmental, Social, and Governance (ESG) reports are central to investment decision-making, yet their length, heterogeneous content, and lack of stan

Exploiting LLM-as-a-Judge Disposition on Free Text Legal QA via Prompt Optimization

arXiv:2604.20726v1 Announce Type: cross Abstract: This work explores the role of prompt design and judge selection in LLM-as-a-Judge evaluations of free text legal question answering. We examine wheth

From Actions to Understanding: Conformal Interpretability of Temporal Concepts in LLM Agents

arXiv:2604.19775v1 Announce Type: new Abstract: Large Language Models (LLMs) are increasingly deployed as autonomous agents capable of reasoning, planning, and acting within interactive environments.

agentsarxiv-cs-ai

Taint-Style Vulnerability Detection and Confirmation for Node.js Packages Using LLM Agent Reasoning

arXiv:2604.20179v1 Announce Type: cross Abstract: The rapidly evolving Node.js ecosystem currently includes millions of packages and is a critical part of modern software supply chains, making vulnera

Transparent Screening for LLM Inference and Training Impacts

arXiv:2604.19757v1 Announce Type: cross Abstract: This paper presents a transparent screening framework for estimating inference and training impacts of current large language models under limited obs

Beyond Indistinguishability: Measuring Extraction Risk in LLM APIs

arXiv:2604.18697v1 Announce Type: cross Abstract: Indistinguishability properties such as differential privacy bounds or low empirically measured membership inference are widely treated as proxies to

DASH-KV: Accelerating Long-Context LLM Inference via Asymmetric KV Cache Hashing

arXiv:2604.19351v1 Announce Type: new Abstract: The quadratic computational complexity of the standard attention mechanism constitutes a fundamental bottleneck for large language models in long-contex

Local Ai

Detoxification for LLM: From Dataset Itself

arXiv:2604.19124v1 Announce Type: new Abstract: Existing detoxification methods for large language models mainly focus on post-training stage or inference time, while few tackle the source of toxicity

local-aiarxiv-cs-cl

Dynamic Model Routing and Cascading for Efficient LLM Inference: A Survey

arXiv:2603.04445v2 Announce Type: replace-cross Abstract: The rapid growth of large language models (LLMs) with diverse capabilities, costs, and domains has created a critical need for intelligent mod

Evaluating Answer Leakage Robustness of LLM Tutors against Adversarial Student Attacks

arXiv:2604.18660v1 Announce Type: cross Abstract: Large Language Models (LLMs) are increasingly used in education, yet their default helpfulness often conflicts with pedagogical principles. Prior work

EVPO: Explained Variance Policy Optimization for Adaptive Critic Utilization in LLM Post-Training

arXiv:2604.19485v1 Announce Type: cross Abstract: Reinforcement learning (RL) for LLM post-training faces a fundamental design choice: whether to use a learned critic as a baseline for policy optimiza

safetyarxiv-cs-ai

See2Refine: Vision-Language Feedback Improves LLM-Based eHMI Action Designers

arXiv:2602.02063v2 Announce Type: replace-cross Abstract: Automated vehicles lack natural communication channels with other road users, making external Human-Machine Interfaces (eHMIs) essential for c

Understanding LLM Performance Degradation in Multi-Instance Processing: The Roles of Instance Count and Context Length

arXiv:2603.22608v2 Announce Type: replace Abstract: Users often rely on Large Language Models (LLMs) for processing multiple documents or performing analysis over a number of instances. For example, a

Before You Interpret the Profile: Validity Scaling for LLM Metacognitive Self-Report

arXiv:2604.17707v1 Announce Type: new Abstract: Clinical personality assessment screens response validity before interpreting substantive scales. LLM evaluation does not. We apply the validity scaling

21 Apr 2026

Beyond URLs: Metadata Diversity and Position for Efficient LLM Pretraining

arXiv:2511.21613v2 Announce Type: replace Abstract: Incorporating metadata in Large Language Models (LLMs) pretraining has recently emerged as a promising approach to accelerate training. However prio