惯性聚合 高效追踪和阅读你感兴趣的博客、新闻、科技资讯
阅读原文 在惯性聚合中打开

推荐订阅源

WordPress大学
WordPress大学
Stack Overflow Blog
Stack Overflow Blog
MongoDB | Blog
MongoDB | Blog
小众软件
小众软件
U
Unit 42
S
SegmentFault 最新的问题
A
About on SuperTechFans
T
Tailwind CSS Blog
Hugging Face - Blog
Hugging Face - Blog
H
Help Net Security
freeCodeCamp Programming Tutorials: Python, JavaScript, Git & More
Recorded Future
Recorded Future
V
Visual Studio Blog
G
Google Developers Blog
The GitHub Blog
The GitHub Blog
Cyber Security Advisories - MS-ISAC
Cyber Security Advisories - MS-ISAC
I
InfoQ
奇客Solidot–传递最新科技情报
奇客Solidot–传递最新科技情报
Y
Y Combinator Blog
博客园 - 司徒正美
量子位
美团技术团队
云风的 BLOG
云风的 BLOG
B
Blog RSS Feed
酷 壳 – CoolShell
酷 壳 – CoolShell
D
Docker
J
Java Code Geeks
B
Blog
L
LangChain Blog
博客园 - 叶小钗
雷峰网
雷峰网
博客园_首页
F
Fortinet All Blogs
Recent Announcements
Recent Announcements
Google DeepMind News
Google DeepMind News
The Cloudflare Blog
Engineering at Meta
Engineering at Meta
有赞技术团队
有赞技术团队
H
Hackread – Cybersecurity News, Data Breaches, AI and More
GbyAI
GbyAI
Blog — PlanetScale
Blog — PlanetScale
Microsoft Azure Blog
Microsoft Azure Blog
阮一峰的网络日志
阮一峰的网络日志
P
Proofpoint News Feed
博客园 - 聂微东
腾讯CDC
T
The Blog of Author Tim Ferriss
罗磊的独立博客
CTFtime.org: upcoming CTF events
CTFtime.org: upcoming CTF events
博客园 - 三生石上(FineUI控件)

cs.CL updates on arXiv.org

Indexing Multimodal Language Models for Large-scale Image Retrieval SpatialEvo: Self-Evolving Spatial Intelligence via Deterministic Geometric Environments PersonaVLM: Long-Term Personalized Multimodal LLMs MedRCube: A Multidimensional Framework for Fine-Grained and In-Depth Evaluation of MLLMs in Medical Imaging Who Gets Flagged? The Pluralistic Evaluation Gap in AI Content Watermarking Addressing Overthinking in Large Vision-Language Models via Gated Perception-Reasoning Optimization VLMs Need Words: Vision Language Models Ignore Visual Detail In Favor of Semantic Anchors (How) Learning Rates Regulate Catastrophic Overtraining Parameter Importance is Not Static: Evolving Parameter Isolation for Supervised Fine-Tuning $π$-Play: Multi-Agent Self-Play via Privileged Self-Distillation without External Data A Domain-Specific Language for LLM-Driven Trigger Generation in Multimodal Data Collection The Consciousness Cluster: Emergent preferences of Models that Claim to be Conscious KMMMU: Evaluation of Massive Multi-discipline Multimodal Understanding in Korean Language and Context Dental-TriageBench: Benchmarking Multimodal Reasoning for Hierarchical Dental Triage Detection Without Correction: A Robust Asymmetry in Activation-Based Hallucination Probing Better and Worse with Scale: How Contextual Entrainment Diverges with Model Size C2: Scalable Rubric-Augmented Reward Modeling from Binary Preferences Calibrated Speculative Decoding: Frequency-Guided Candidate Selection for Efficient Inference A Multi-Model Approach to English-Bangla Sentiment Classification of Government Mobile Banking App Reviews Mathematical Reasoning Enhanced LLM for Formula Derivation: A Case Study on Fiber NLI Modellin Red Skills or Blue Skills? A Dive Into Skills Published on ClawHub Can Large Language Models Reliably Extract Physiology Index Values from Coronary Angiography Reports? IWLV-Ramayana: A Sarga-Aligned Parallel Corpus of Valmiki's Ramayana Across Indian Languages Unleashing Implicit Rewards: Prefix-Value Learning for Distribution-Level Optimization Evaluating the Evaluator: Problems with SemEval-2020 Task 1 for Lexical Semantic Change Detection From $P(y|x)$ to $P(y)$: Investigating Reinforcement Learning in Pre-train Space UI-Zoomer: Uncertainty-Driven Adaptive Zoom-In for GUI Grounding Reward Design for Physical Reasoning in Vision-Language Models Chain of Uncertain Rewards with Large Language Models for Reinforcement Learning MERRIN: A Benchmark for Multimodal Evidence Retrieval and Reasoning in Noisy Web Environments L2D-Clinical: Learning to Defer for Adaptive Model Selection in Clinical Text Classification Hessian-Enhanced Token Attribution (HETA): Interpreting Autoregressive LLMs InfiniteScienceGym: An Unbounded, Procedurally-Generated Benchmark for Scientific Analysis Toward Generalized Cross-Lingual Hateful Language Detection with Web-Scale Data and Ensemble LLM Annotations Self-Calibrating Language Models via Test-Time Discriminative Distillation HumorGen: Cognitive Synergy for Humor Generation in Large Language Models via Persona-Based Distillation Claim2Vec: Embedding Fact-Check Claims for Multilingual Similarity and Clustering Spoiler Alert: Narrative Forecasting as a Metric for Tension in LLM Storytelling Human vs. Machine Deception: Distinguishing AI-Generated and Human-Written Fake News Using Ensemble Learning Weird Generalization is Weirdly Brittle Mirroring Minds: Asymmetric Linguistic Accommodation and Diagnostic Identity in ADHD and Autism Reddit Communities Why Supervised Fine-Tuning Fails to Learn: A Systematic Study of Incomplete Learning in Large Language Models Reason Only When Needed: Efficient Generative Reward Modeling via Model-Internal Uncertainty SEPTQ: A Simple and Effective Post-Training Quantization Paradigm for Large Language Models Who Wrote This Line? Evaluating the Detection of LLM-Generated Classical Chinese Poetry Training-Free Cross-Lingual Dysarthria Severity Assessment via Phonological Subspace Analysis in Self-Supervised Speech Representations Simulating Organized Group Behavior: New Framework, Benchmark, and Analysis ODUTQA-MDC: A Task for Open-Domain Underspecified Tabular QA with Multi-turn Dialogue-based Clarification Nationality encoding in language model hidden states: Probing culturally differentiated representations in persona-conditioned academic text Relational Probing: LM-to-Graph Adaptation for Financial Prediction CodeComp: Structural KV Cache Compression for Agentic Coding FAITH: Factuality Alignment through Integrating Trustworthiness and Honestness Comparative Analysis of Large Language Models in Healthcare Adaptive Multi-Expert Reasoning via Difficulty-Aware Routing and Uncertainty-Guided Aggregation A Structured Clustering Approach for Inducing Media Narratives NameBERT: Scaling Name-Based Nationality Classification with LLM-Augmented Open Academic Data LASQ: A Low-resource Aspect-based Sentiment Quadruple Extraction Dataset BLUEmed: Retrieval-Augmented Multi-Agent Debate for Clinical Error Detection Turing or Cantor: That is the Question NOSE: Neural Olfactory-Semantic Embedding with Tri-Modal Orthogonal Contrastive Learning Instruction Data Selection via Answer Divergence EviCare: Enhancing Diagnosis Prediction with Deep Model-Guided Evidence for In-Context Reasoning Dynamic Adaptive Attention and Supervised Contrastive Learning: A Novel Hybrid Framework for Text Sentiment Classification Structure-Grounded Knowledge Retrieval via Code Dependencies for Multi-Step Data Reasoning Knowing What to Stress: A Discourse-Conditioned Text-to-Speech Benchmark HeceTokenizer: A Syllable-Based Tokenization Approach for Turkish Retrieval BlasBench: An Open Benchmark for Irish Speech Recognition OccuBench: Evaluating AI Agents on Real-World Professional Tasks via Language Environment Simulation How Robust Are Large Language Models for Clinical Numeracy? An Empirical Study on Numerical Reasoning Abilities in Clinical Contexts Evaluating Memory Capability in Continuous Lifelog Scenario Polyglot Teachers: Evaluating Language Models for Multilingual Synthetic Data Generation Hidden Measurement Error in LLM Pipelines Distorts Annotation, Evaluation, and Benchmarking LangFlow: Continuous Diffusion Rivals Discrete in Language Modeling ProGAL-VLA: Grounded Alignment through Prospective Reasoning in Vision-Language-Action Models Reproduction Beyond Benchmarks: ConstBERT and ColBERT-v2 Across Backends and Query Distributions Hijacking Text Heritage: Hiding the Human Signature through Homoglyphic Substitution SpectralLoRA: Is Low-Frequency Structure Sufficient for LoRA Adaptation? A Spectral Analysis of Weight Updates Revisiting Compositionality in Dual-Encoder Vision-Language Models: The Role of Inference GenProve: Learning to Generate Text with Fine-Grained Provenance ChemPro: A Progressive Chemistry Benchmark for Large Language Models MEDSYN: Benchmarking Multi-EviDence SYNthesis in Complex Clinical Cases for Multimodal Large Language Models PICon: A Multi-Turn Interrogation Framework for Evaluating Persona Agent Consistency CAMO: A Class-Aware Minority-Optimized Ensemble for Robust Language Model Evaluation on Imbalanced Data BadGraph: A Backdoor Attack Against Latent Diffusion Model for Text-Guided Graph Generation Pay Less Attention to Function Words for Free Robustness of Vision-Language Models From Speech-to-Spatial: Grounding Utterances on A Live Shared View with Augmented Reality MerNav: A Highly Generalizable Memory-Execute-Review Framework for Zero-Shot Object Goal Navigation Rethinking LLM Watermark Detection in Black-Box Settings: A Non-Intrusive Third-Party Framework Not All Denoising Steps Are Equal: Model Scheduling for Faster Masked Diffusion Language Models SODA: Semi On-Policy Black-Box Distillation for Large Language Models C-ReD: A Comprehensive Chinese Benchmark for AI-Generated Text Detection Derived from Real-World Prompts Evaluating Cooperation in LLM Social Groups through Elected Leadership A Triadic Suffix Tokenization Scheme for Numerical Reasoning Policy Split: Incentivizing Dual-Mode Exploration in LLM Reinforcement with Dual-Mode Entropy Regularization METER: Evaluating Multi-Level Contextual Causal Reasoning in Large Language Models Anthropogenic Regional Adaptation in Multimodal Vision-Language Model METRO: Towards Strategy Induction from Expert Dialogue Transcripts for Non-collaborative Dialogues Towards Proactive Information Probing: Customer Service Chatbots Harvesting Value from Conversation A Systematic Analysis of the Impact of Persona Steering on LLM Capabilities Linear Representations of Hierarchical Concepts in Language Models
Utility-Diversity Aware Online Batch Selection for LLM Supervised Fine-tuning
Heming Zou, Yixiu Mao, Yun Qu, Qi Wang, Xiangyang Ji · 2025-10-19 · via cs.CL updates on arXiv.org

Supervised fine-tuning (SFT) is a commonly used technique to adapt large language models (LLMs) to downstream tasks. In practice, SFT on a full dataset is computationally expensive and sometimes suffers from overfitting or bias amplification. This facilitates the rise of data curation in SFT, which prioritizes the most valuable data to optimze. This work studies the online batch selection family that dynamically scores and filters samples during the training process. However, existing popular methods often (i) rely merely on the utility of data to select a subset while neglecting other crucial factors like diversity, (ii) rely on external resources such as reference models or validation sets, and (iii) incur extra training time over full-dataset training. To address these limitations, this work develops UDS (Utility-Diversity Sampling), a framework for efficient online batch selection in SFT. UDS leverages the nuclear norm of the logits matrix to capture both data utility and intra-sample diversity, while estimating inter-sample diversity through efficient low-dimensional embedding comparisons with a lightweight memory buffer of historical samples. Such a design eliminates the need for external resources and unnecessary backpropagation, securing computational efficiency. Experiments on multiple benchmarks demonstrate that UDS consistently outperforms state-of-the-art online batch selection methods under varying data budgets, and significantly reduces training time compared to full-dataset fine-tuning. Code is available at https://github.com/gfyddha/UDS.