This paper introduces DraDDP, the first publicly available English multimodal dataset for multi-party dialogue discourse parsing, based on American TV dramas. It contains 495 dialogue segments and demonstrates the value of multimodal information in capturing dialogue structures and relation types.
Modelos de linguagem de difusão discreta (dLLMs) aceleram a geração de texto, mas a decodificação paralela degrada a qualidade ao desconsiderar a dependência entre tokens. DEMASK propõe um preditor leve que estima influências condicionais para guiar o desmascaramento simultâneo, comprovadamente melhorando a qualidade. A técnica resulta em um ganho de velocidade de 1.7 a 2.2x, mantendo ou superando o desempenho.
The paper proposes a novel Multi-Granularity Reasoning Network (MGRN) for Natural Language Inference (NLI). It addresses the limitations of existing transformer-based models by leveraging hierarchical semantic features to capture complex interactions for effective reasoning.
This paper introduces a non-autoregressive scoring method for efficient punctuation restoration in streaming Automatic Speech Recognition (ASR) systems. It compares punctuation insertion hypotheses against a no-insertion baseline using a bounded K-subword-token lookahead, outperforming existing prompt-based methods.
This research explores a novel approach to solving math word problems by integrating the power of language models with the precision of symbolic solvers. The method aims to leverage both natural language understanding and formal mathematical reasoning to achieve robust solutions.
This content explains how AI automation can streamline literature reviews, turning PDF data extraction into a simplified, less error-prone process. It highlights the importance of an iterative refinement loop and introduces the open-source GROBID library for structured academic data extraction.
This content describes how AI can automate the complex task of finding inconsistencies across multiple witness statements for legal cross-examination. The method involves moving from individual statement summarization to a unified comparative analysis through entity and event alignment.
This post explains how to integrate Amazon Quick with AWS services using Amazon Bedrock AgentCore Runtime's Model Context Protocol (MCP) support. It demonstrates creating a conversational AI assistant that translates natural language into AWS CLI commands, streamlining operations.
ArcDeck is a multi-agent AI framework that generates slides from academic papers by explicitly modeling the paper's logical flow and narrative structure. It uses a discourse tree and iterative agent-based refinement to ensure coherence, demonstrating significant improvements in generated presentations.
OThink-SRR1 is a framework that enhances LLMs with an iterative Search-Refine-Reason process trained via reinforcement learning. It addresses RAG's challenges by distilling relevant facts from retrieved documents, improving efficiency and accuracy in complex multi-hop QA.
Large language models struggle with complex long-context reasoning tasks despite supporting extensive inputs. ProxyCoT is a novel training framework designed to transfer reasoning capabilities from short proxy contexts to full long contexts, outperforming strong baselines.
A systematic inspection of extsf{FOLIO} and extsf{MALLS} validation splits revealed high rates of incorrect FOL formalizations and ambiguous NL sentences, distorting AI model evaluation. The authors developed and released corrected ground truths for these datasets, demonstrating how annotation errors impact the evaluation of state-of-the-art LLMs.
This article describes a problem encountered when deploying a new AI encoder, which, despite significantly improving concept identification, broke every response. It details the role of the "Dispatcher" in the Origin system, acting as an intermediary between the encoder and the response, processing concept activations to determine appropriate actions.
This guide details an end-to-end Machine Learning pipeline for email spam classification. It compares Naive Bayes and RoBERTa models, visualizes with Streamlit, and deploys using Docker.
The paper introduces ASL-MTP, a new benchmark dataset for analyzing how well sign language models capture linguistic phenomena and utilize multi-articulator cues. It uses this dataset to conduct a targeted linguistic analysis of a state-of-the-art ASL-to-English translation model.
This research explores how humans communicate with limited vocabularies, comparing their strategies to computational sampling algorithms powered by large language models. The study reveals that human language production under constraint often mirrors greedy sampling, although more skilled individuals exhibit non-greedy revision behaviors.
This research investigates the balance between fluency and faithfulness in literary translation, comparing human, Google Translate, and TranslateGemma performance across 106 novels in 16 source languages. It reveals a consistent negative correlation between fluency and faithfulness, particularly for human and Google Translate, and indicates that segment length significantly impacts automatic evaluation.
This research introduces LIFT, a learnability-informed fine-tuning algorithm designed to enhance the reasoning capabilities of diffusion language models. LIFT addresses the shortcomings of standard SFT by adaptively learning tokens based on their difficulty and available context during different diffusion time steps, showing improved performance over existing baselines.
This article addresses the productivity bottleneck caused by manually deciphering technician voice notes, proposing AI as a solution to transform field recordings into professional summaries. It outlines a methodology, the 'Actionable Framework: The 3-Part Jargon List,' to train AI to categorize specific information from unstructured audio.
Este conteúdo apresenta uma arquitetura RAG (Retrieval Augmented Generation) controlada que utiliza perguntas de competência como planos executáveis. O objetivo é aplicar essa metodologia para a criação de narrativas no campo do patrimônio cultural.