Ep. 261 - Part 1 - June 11, 2024

ArXiv NLP research for Tuesday, June 11, 2024. 00:20: A Non-autoregressive Generation Framework for End-to-End Simultaneous Speech-to-Any Translation 01:41: Post-Hoc Answer Attribution for Grounded and Trustworthy Long Document Comprehension: Task, Insights, and Challenges 02:32: A Probabilistic Framework for LLM Hallucination Detection via Belief Tree Propagation 04:08: Evolving Subnetwork Training for Large Language Models 05:31: Missingness-resilient Video-enhanced Multimodal Disfluency Detection 06:37: Mitigating Boundary Ambiguity and Inherent Bias for Text Classification in the Era of Large Language Models 08:14: Crayon: Customized On-Device LLM via Instant Adapter Blending and Edge-Server Hybrid Inference 09:33: Delving into ChatGPT usage in academic writing through excess vocabulary 10:53: Paying More Attention to Source Context: Mitigating Unfaithful Translations from Large Language Model 12:12: CoEvol: Constructing Better Responses for Instruction Finetuning through Multi-Agent Cooperation 13:26: Effectively Compress KV Heads for LLM 15:00: Benchmarking Trustworthiness of Multimodal Large Language Models: A Comprehensive Study 16:54: Reading Miscue Detection in Primary School through Automatic Speech Recognition 18:09: HalluDial: A Large-Scale Benchmark for Automatic Dialogue-Level Hallucination Evaluation 20:01: DARA: Decomposition-Alignment-Reasoning Autonomous Language Agent for Question Answering over Knowledge Graphs 21:15: Efficiently Exploring Large Language Models for Document-Level Machine Translation with In-context Learning 22:35: Advancing Tool-Augmented Large Language Models: Integrating Insights from Errors in Inference Trees 24:42: Translating speech with just images 25:35: Never Miss A Beat: An Efficient Recipe for Context Window Extension of Large Language Models with Consistent "Middle" Enhancement 26:51: Teaching Language Models to Self-Improve by Learning from Language Feedback 28:25: Merging Improves Self-Critique Against Jailbreak Attacks 29:18: Towards Human-AI Collaboration in Healthcare: Guided Deferral Systems with Large Language Models 30:11: Improving Autoformalization using Type Checking 31:37: Improving Commonsense Bias Classification by Mitigating the Influence of Demographic Terms 33:19: Decipherment-Aware Multilingual Learning in Jointly Trained Language Models 34:20: DUAL-REFLECT: Enhancing Large Language Models for Reflective Translation through Dual Learning Feedback Mechanisms 35:20: On the Hallucination in Simultaneous Machine Translation 36:07: MBBQ: A Dataset for Cross-Lingual Comparison of Stereotypes in Generative LLMs 37:42: Scholarly Question Answering using Large Language Models in the NFDI4DataScience Gateway