Search: “arxiv-cs-cl”

Decomposing and Reducing Hidden Measurement Error in LLM Evaluation Pipelines

arXiv:2604.11581v1 Announce Type: new Abstract: LLM evaluations drive which models get deployed, which safety standards get adopted, and which research conclusions get published. Yet these scores carr

Content type

All Blog X Post Paper YouTube Reddit GitHub

Clear filters

Research

DeCoVec: Building Decoding Space based Task Vector for Large Language Models via In-Context Learning

DGX agent

arXiv:2604.11129v1 Announce Type: new Abstract: Task vectors, representing directions in model or activation spaces that encode task-specific behaviors, have emerged as a promising tool for steering l

researcharxiv-cs-cl

14 Apr 2026

Defending against Backdoor Attacks via Module Switching

arXiv:2504.05902v2 Announce Type: replace-cross Abstract: Backdoor attacks pose a serious threat to deep neural networks (DNNs), allowing adversaries to implant triggers for hidden behaviors in infere

Dialectic-Med: Mitigating Diagnostic Hallucinations via Counterfactual Adversarial Multi-Agent Debate

arXiv:2604.11258v1 Announce Type: new Abstract: Multimodal Large Language Models (MLLMs) in healthcare suffer from severe confirmation bias, often hallucinating visual details to support initial, pote

Local Ai

Different types of syntactic agreement recruit the same units within large language models

arXiv:2512.03676v2 Announce Type: replace Abstract: Large language models (LLMs) can reliably distinguish grammatical from ungrammatical sentences, but how grammatical knowledge is represented within

local-aiarxiv-cs-cl

Discourse Coherence and Response-Guided Context Rewriting for Multi-Party Dialogue Generation

arXiv:2604.06784v2 Announce Type: replace Abstract: Previous research on multi-party dialogue generation has predominantly leveraged structural information inherent in dialogues to directly inform the

Doc-PP: Document Policy Preservation Benchmark for Large Vision-Language Models

arXiv:2601.03926v2 Announce Type: replace Abstract: The deployment of Large Vision-Language Models (LVLMs) for real-world document question answering is often constrained by dynamic, user-defined poli

DuET: Dual Execution for Test Output Prediction with Generated Code and Pseudocode

arXiv:2604.11514v1 Announce Type: cross Abstract: This work addresses test output prediction, a key challenge in test case generation. To improve the reliability of predicted outputs by LLMs, prior ap

Dynamic Adaptive Attention and Supervised Contrastive Learning: A Novel Hybrid Framework for Text Sentiment Classification

arXiv:2604.10459v1 Announce Type: new Abstract: The exponential growth of user-generated movie reviews on digital platforms has made accurate text sentiment classification a cornerstone task in natura

EEPO: Exploration-Enhanced Policy Optimization via Sample-Then-Forget

arXiv:2510.05837v2 Announce Type: replace Abstract: Balancing exploration and exploitation remains a central challenge in reinforcement learning with verifiable rewards (RLVR) for large language model

Eliciting Medical Reasoning with Knowledge-enhanced Data Synthesis: A Semi-Supervised Reinforcement Learning Approach

arXiv:2604.11547v1 Announce Type: cross Abstract: While large language models hold promise for complex medical applications, their development is hindered by the scarcity of high-quality reasoning dat

End-to-end Contrastive Language-Speech Pretraining Model For Long-form Spoken Question Answering

arXiv:2511.09282v3 Announce Type: replace-cross Abstract: Significant progress has been made in spoken question answering (SQA) in recent years. However, many existing methods, including large audio l

Local Ai

Enhancing Multilingual RAG Systems with Debiased Language Preference-Guided Query Fusion

arXiv:2601.02956v3 Announce Type: replace Abstract: Multilingual Retrieval-Augmented Generation (mRAG) systems often exhibit a perceived preference for high-resource languages, particularly English, r

local-aiarxiv-cs-cl

Evaluating Memory Capability in Continuous Lifelog Scenario

arXiv:2604.11182v1 Announce Type: new Abstract: Nowadays, wearable devices can continuously lifelog ambient conversations, creating substantial opportunities for memory systems. However, existing benc

Evaluating Small Open LLMs for Medical Question Answering: A Practical Framework

arXiv:2604.10535v1 Announce Type: cross Abstract: Incorporating large language models (LLMs) in medical question answering demands more than high average accuracy: a model that returns substantively d

Tutorials

EviCare: Enhancing Diagnosis Prediction with Deep Model-Guided Evidence for In-Context Reasoning

arXiv:2604.10455v1 Announce Type: new Abstract: Recent advances in large language models (LLMs) have enabled promising progress in diagnosis prediction from electronic health records (EHRs). However,

tutorialsarxiv-cs-cl

EvoDiagram: Agentic Editable Diagram Creation via Design Expertise Evolution

arXiv:2604.09568v1 Announce Type: cross Abstract: High-fidelity diagram creation requires the complex orchestration of semantic topology, visual styling, and spatial layout, posing a significant chall

Expect the Unexpected? Testing the Surprisal of Salient Entities

arXiv:2604.10724v1 Announce Type: new Abstract: Previous work examining the Uniform Information Density (UID) hypothesis has shown that while information as measured by surprisal metrics is distribute

Exploring Knowledge Purification in Multi-Teacher Knowledge Distillation for LLMs

arXiv:2602.01064v2 Announce Type: replace Abstract: Knowledge distillation has emerged as a pivotal technique for transferring knowledge from stronger large language models (LLMs) to smaller, more eff

FAITH: Factuality Alignment through Integrating Trustworthiness and Honestness

arXiv:2604.10189v1 Announce Type: new Abstract: Large Language Models (LLMs) can generate factually inaccurate content even if they have corresponding knowledge, which critically undermines their reli

Tutorials

Find Your Optimal Teacher: Personalized Data Synthesis via Router-Guided Multi-Teacher Distillation

arXiv:2510.10925v2 Announce Type: replace-cross Abstract: Training student models on synthetic data generated by strong teacher models is a promising way to distilling the capabilities of teachers. Ho

tutorialsarxiv-cs-cl

FlashMem: Distilling Intrinsic Latent Memory via Computation Reuse

arXiv:2601.05505v2 Announce Type: replace Abstract: The stateless architecture of Large Language Models inherently lacks the mechanism to preserve dynamic context, compelling agents to redundantly rep

From Perception to Autonomous Computational Modeling: A Multi-Agent Approach

arXiv:2604.06788v2 Announce Type: replace-cross Abstract: We present a solver-agnostic framework in which coordinated large language model (LLM) agents autonomously execute the complete computational

From Speech-to-Spatial: Grounding Utterances on A Live Shared View with Augmented Reality

arXiv:2602.03059v2 Announce Type: replace-cross Abstract: We introduce Speech-to-Spatial, a referent disambiguation framework that converts verbal remote-assistance instructions into spatially grounde

Generation-Augmented Generation: A Plug-and-Play Framework for Private Knowledge Injection in Large Language Models

arXiv:2601.08209v3 Announce Type: replace Abstract: In domains such as materials science, biomedicine, and finance, high-stakes deployment of large language models (LLMs) requires injecting private, d

GenProve: Learning to Generate Text with Fine-Grained Provenance

arXiv:2601.04932v2 Announce Type: replace Abstract: Large language models (LLM) often hallucinate, and while adding citations is a common solution, it is frequently insufficient for accountability as

Geometry-Aware Localized Watermarking for Copyright Protection in Embedding-as-a-Service

arXiv:2604.11344v1 Announce Type: cross Abstract: Embedding-as-a-Service (EaaS) has become an important semantic infrastructure for natural language and multimedia applications, but it is highly vulne

Applications

Head-wise Modality Specialization within MLLMs for Robust Fake News Detection under Missing Modality

arXiv:2604.09711v1 Announce Type: cross Abstract: Multimodal fake news detection (MFND) aims to verify news credibility by jointly exploiting textual and visual evidence. However, real-world news diss

applicationsarxiv-cs-cl

HeceTokenizer: A Syllable-Based Tokenization Approach for Turkish Retrieval

arXiv:2604.10665v1 Announce Type: new Abstract: HeceTokenizer is a syllable-based tokenizer for Turkish that exploits the deterministic six-pattern phonological structure of the language to construct

Hidden Failures in Robustness: Why Supervised Uncertainty Quantification Needs Better Evaluation

arXiv:2604.11662v1 Announce Type: new Abstract: Recent work has shown that the hidden states of large language models contain signals useful for uncertainty estimation and hallucination detection, mot

HiEdit: Lifelong Model Editing with Hierarchical Reinforcement Learning

arXiv:2604.11214v1 Announce Type: new Abstract: Lifelong model editing (LME) aims to sequentially rectify outdated or inaccurate knowledge in deployed LLMs while minimizing side effects on unrelated i

Tutorials

Hierarchical Textual Knowledge for Enhanced Image Clustering

arXiv:2604.11144v1 Announce Type: cross Abstract: Image clustering aims to group images in an unsupervised fashion. Traditional methods focus on knowledge from visual space, making it difficult to dis

tutorialsarxiv-cs-cl

Hijacking Text Heritage: Hiding the Human Signature through Homoglyphic Substitution

arXiv:2604.10271v1 Announce Type: cross Abstract: In what way could a data breach involving government-issued IDs such as passports, driver's licenses, etc., rival a random voluntary disclosure on a n

HistLens: Mapping Idea Change across Concepts and Corpora

arXiv:2604.11749v1 Announce Type: new Abstract: Language change both reflects and shapes social processes, and the semantic evolution of foundational concepts provides a measurable trace of historical

How Robust Are Large Language Models for Clinical Numeracy? An Empirical Study on Numerical Reasoning Abilities in Clinical Contexts

arXiv:2604.11133v1 Announce Type: new Abstract: Large Language Models (LLMs) are increasingly being explored for clinical question answering and decision support, yet safe deployment critically requir

How You Ask Matters! Adaptive RAG Robustness to Query Variations

arXiv:2604.10745v1 Announce Type: new Abstract: Adaptive Retrieval-Augmented Generation (RAG) promises accuracy and efficiency by dynamically triggering retrieval only when needed and is widely used i

Agents

HTAA: Enhancing LLM Planning via Hybrid Toolset Agentization & Adaptation

arXiv:2604.10917v1 Announce Type: new Abstract: Enabling large language models to scale and reliably use hundreds of tools is critical for real-world applications, yet challenging due to the inefficie

agentsarxiv-cs-cl

Human vs. Machine Deception: Distinguishing AI-Generated and Human-Written Fake News Using Ensemble Learning

arXiv:2604.09960v1 Announce Type: new Abstract: The rapid adoption of large language models has introduced a new class of AI-generated fake news that coexists with traditional human-written misinforma

HumorGen: Cognitive Synergy for Humor Generation in Large Language Models via Persona-Based Distillation

arXiv:2604.09629v1 Announce Type: new Abstract: Humor generation poses a significant challenge for Large Language Models (LLMs), because their standard training objective - predicting the most likely

HyperGraphPro: Progress-Aware Reinforcement Learning for Structure-Guided Hypergraph RAG

arXiv:2601.17755v2 Announce Type: replace Abstract: Graph Retrieval-Augmented Generation (GraphRAG) has emerged as a promising paradigm that organizes external knowledge into structured graphs of enti

Improving LLM Unlearning Robustness via Random Perturbations

arXiv:2501.19202v5 Announce Type: replace Abstract: Here, we show that current LLM unlearning methods inherently reduce models' robustness, causing them to misbehave even when a single non-adversarial

Infusing Theory of Mind into Socially Intelligent LLM Agents

arXiv:2509.22887v2 Announce Type: replace Abstract: Theory of Mind (ToM)-an understanding of the mental states of others-is a key aspect of human social intelligence, yet, chatbots and LLM-based socia

Instruction Data Selection via Answer Divergence

arXiv:2604.10448v1 Announce Type: new Abstract: Instruction tuning relies on large instruction-response corpora whose quality and composition strongly affect downstream performance. We propose Answer

Judge Like Human Examiners: A Weighted Importance Multi-Point Evaluation Framework for Generative Tasks with Long-form Answers

arXiv:2604.11246v1 Announce Type: new Abstract: Evaluating the quality of model responses remains challenging in generative tasks with long-form answers, as the expected answers usually contain multip

K-Way Energy Probes for Metacognition Reduce to Softmax in Discriminative Predictive Coding Networks

arXiv:2604.11011v1 Announce Type: cross Abstract: We present this as a negative result with an explanatory mechanism, not as a formal upper bound. Predictive coding networks (PCNs) admit a K-way energ

KCS: Diversify Multi-hop Question Generation with Knowledge Composition Sampling

arXiv:2508.20567v2 Announce Type: replace Abstract: Multi-hop question answering faces substantial challenges due to data sparsity, which increases the likelihood of language models learning spurious

Knowing What to Stress: A Discourse-Conditioned Text-to-Speech Benchmark

arXiv:2604.10580v1 Announce Type: new Abstract: Spoken meaning often depends not only on what is said, but also on which word is emphasized. The same sentence can convey correction, contrast, or clari

ks-pret-5m: a 5 million word, 12 million token kashmiri pretraining dataset

arXiv:2604.11066v1 Announce Type: new Abstract: We present KS-PRET-5M, the largest publicly available pretraining dataset for the Kashmiri language, comprising 5,090,244 (5.09M) words, 27,692,959 (27.