Image de couverture de livre

What Works for 'Lost-in-the-Middle' in LLMs? A Study on GM-Extract and Mitigations

Fuente: arXiv

Enregistré dans:

Détails bibliographiques
Auteurs principaux:	Gupte, Mihir, Dixit, Eshan, Tayyab, Muhammad, Adiththan, Arun
Format:	Preprint
Publié:	2025
Sujets:	Computation and Language Artificial Intelligence
Accès en ligne:	Acceder al recurso
Tags:	Ajouter un tag Pas de tags, Soyez le premier à ajouter un tag!

Documents similaires

Towards Autoformalization of LLM-generated Outputs for Requirement Verification
par: Gupte, Mihir, et autres
Publié: (2025)

Is Implicit Knowledge Enough for LLMs? A RAG Approach for Tree-based Structures
par: Gupte, Mihir, et autres
Publié: (2025)

Lost-in-the-Middle in Long-Text Generation: Synthetic Dataset, Evaluation Framework, and Mitigation
par: Zhang, Junhao, et autres
Publié: (2025)

What Has Been Lost with Synthetic Evaluation?
par: Gill, Alexander, et autres
Publié: (2025)

Lost in the Middle at Birth: An Exact Theory of Transformer Position Bias
par: Chowdhury, Borun D
Publié: (2026)

Never Lost in the Middle: Mastering Long-Context Question Answering with Position-Agnostic Decompositional Training
par: He, Junqing, et autres
Publié: (2023)

Lost in Stories: Consistency Bugs in Long Story Generation by LLMs
par: Li, Junjie, et autres
Publié: (2026)

The Script Tax: Measuring Tokenization-Driven Efficiency and Latency Disparities in Multilingual Language Models
par: Dixit, Aradhya, et autres
Publié: (2026)

Directional Optimization Asymmetry in Transformers: A Synthetic Stress Test
par: Sahasrabudhe, Mihir
Publié: (2025)

Retracing the Past: LLMs Emit Training Data When They Get Lost
par: Ko, Myeongseob, et autres
Publié: (2025)

Intent Mismatch Causes LLMs to Get Lost in Multi-Turn Conversation
par: Liu, Geng, et autres
Publié: (2026)

Extracting Breast Cancer Phenotypes from Clinical Notes: Comparing LLMs with Classical Ontology Methods
par: Faiz, Abdullah Bin, et autres
Publié: (2026)

What Makes Low-Bit Quantization-Aware Training Work for Reasoning LLMs? A Systematic Study
par: Lv, Keyu, et autres
Publié: (2026)

Lost in Sampling: Assessing Lexical Reachability in LLMs via the Word Coverage Score (WCS)
par: Awad, Samer, et autres
Publié: (2026)

Lost in Translation? A Comparative Study on the Cross-Lingual Transfer of Composite Harms
par: Shukla, Vaibhav, et autres
Publié: (2026)

EDU-NER-2025: Named Entity Recognition in Urdu Educational Texts using XLM-RoBERTa with X (formerly Twitter)
par: Ullah, Fida, et autres
Publié: (2025)

Towards Enhancing Coherence in Extractive Summarization: Dataset and Experiments with LLMs
par: Parmar, Mihir, et autres
Publié: (2024)

Lost in Time: Clock and Calendar Understanding Challenges in Multimodal LLMs
par: Saxena, Rohit, et autres
Publié: (2025)

Not All Documents Are What You Need for Extracting Instruction Tuning Data
par: Zhang, Chi, et autres
Publié: (2025)

Fairness Evaluation and Inference Level Mitigation in LLMs
par: Nadeem, Afrozah, et autres
Publié: (2025)

Quantifying and Mitigating Premature Closure in Frontier LLMs
par: Handler, Rebecca, et autres
Publié: (2026)

Mitigating Lost in Multi-turn Conversation via Curriculum RL with Verifiable Accuracy and Abstention Rewards
par: Li, Ming, et autres
Publié: (2025)

Decompose, Enrich, and Extract! Schema-aware Event Extraction using LLMs
par: Shiri, Fatemeh, et autres
Publié: (2024)

Extract Information from Hybrid Long Documents Leveraging LLMs: A Framework and Dataset
par: Yue, Chongjian, et autres
Publié: (2024)

Steering LLMs for Formal Theorem Proving
par: Kirtania, Shashank, et autres
Publié: (2025)

A Critical Study of What Code-LLMs (Do Not) Learn
par: Anand, Abhinav, et autres
Publié: (2024)

Meta-Thinking in LLMs via Multi-Agent Reinforcement Learning: A Survey
par: Bilal, Ahsan, et autres
Publié: (2025)

Steering Towards Fairness: Mitigating Political Bias in LLMs
par: Nadeem, Afrozah, et autres
Publié: (2025)

Lottery Ticket Adaptation: Mitigating Destructive Interference in LLMs
par: Panda, Ashwinee, et autres
Publié: (2024)

Learning to Refuse: Towards Mitigating Privacy Risks in LLMs
par: Liu, Zhenhua, et autres
Publié: (2024)

Middle-Layer Representation Alignment for Cross-Lingual Transfer in Fine-Tuned LLMs
par: Liu, Danni, et autres
Publié: (2025)

Spanish and LLM Benchmarks: is MMLU Lost in Translation?
par: Plaza, Irene, et autres
Publié: (2024)

Pause-Tuning for Long-Context Comprehension: A Lightweight Approach to LLM Attention Recalibration
par: Begin, James, et autres
Publié: (2025)

Intermittent Semi-Working Mask: A New Masking Paradigm for LLMs
par: Hu, HaoYuan, et autres
Publié: (2024)

Automatic Posology Structuration : What role for LLMs?
par: Bobkova, Natalia, et autres
Publié: (2025)

What Makes Cryptic Crosswords Challenging for LLMs?
par: Sadallah, Abdelrahman, et autres
Publié: (2024)

Step-by-Step Reasoning to Solve Grid Puzzles: Where do LLMs Falter?
par: Tyagi, Nemika, et autres
Publié: (2024)

When Personalization Misleads: Understanding and Mitigating Hallucinations in Personalized LLMs
par: Sun, Zhongxiang, et autres
Publié: (2026)

Enabling and Analyzing How to Efficiently Extract Information from Hybrid Long Documents with LLMs
par: Yue, Chongjian, et autres
Publié: (2023)

Time-Reversal Provides Unsupervised Feedback to LLMs
par: Varun, Yerram, et autres
Publié: (2024)