-
Learning to Predict Situation Hyper-Graphs for Video Question Answering
The SHG-VQA model predicts a situation hyper-graph structure composed of existing actions and relations in the input video. -
Planning by Automatic Prompt Engineering for Large Language Models Agents
The paper proposes a novel method, REPROMPT, for optimizing the step-by-step instructions in the prompt of LLM agents based on the chat history obtained from interactions with... -
SemEval-2016 (Task 3A)
SemEval-2016 (Task 3A) is a sentence-level QA dataset, containing 1.8k/0.2k/0.3k train/dev/test examples. Each example consists of a community question by a user and 10 comments. -
Question Answering through Transfer Learning
Question answering (QA) is a long-standing challenge in NLP, and the community has introduced several paradigms and datasets for the task over the past few years. These... -
KnowIT VQA
A video story question answering dataset containing 24,282 questions about 207 episodes of The Big Bang Theory. -
MARGE: A Pre-trained Sequence-to-Sequence Model for Multi-lingual Paraphrasing
MARGE is a pre-trained sequence-to-sequence model learned with an unsupervised multi-lingual multi-document paraphrasing objective. -
Alquist: The Alexa Prize Socialbot
The Alquist dialogue system is designed to conduct a coherent and engaging conversation on popular topics. -
LogicVista: Multimodal LLM Logical Reasoning Benchmark in Visual Contexts
LogicVista is a comprehensive evaluation benchmark for multimodal large language models (MLLMs) in visual contexts. It assesses the integrated logical reasoning capabilities of... -
ShortcutQA
The authors created a dataset called ShortcutQA, which is a curated dataset generated by their framework for future research. -
RAGSummData
The dataset used in the paper is a collection of dialogues and prompts for training a model to perform retrieval-augmented generation (RAG) based summarization. The dataset is... -
CommonsenseQA and OpenBookQA
CommonsenseQA and OpenBookQA are two of the most widely used commonsense reasoning benchmarks. -
Ontology-based question answering over corporate structured data
Ontology-based question answering over corporate structured data -
ConceptNet 5.5
The ConceptNet 5.5 dataset is an open multilingual graph of general knowledge. -
Neural-Symbolic Commonsense Reasoner with Relation Predictors
A neural-symbolic reasoner for Commonsense Knowledge Graphs (CKGs) that leverages a relation prediction module and weak unification. -
Existing ACQ datasets
A few existing datasets for asking clarification questions -
FLM-HotpotQA
A dataset for pragmatic evaluation of clarifying questions and fact-level masking -
Generative Agents framework
Generative Agents framework by Park et al., aimed at enhancing the efficient retrieval of key events for general-purpose LLM agents.