arXiv Comp. Linguistics๐Ÿ“š · @arxiv_cl
222 followers · 3987 posts · Server creative.ai

๐Ÿ“ Exploring Large Language Models for Communication Games: An Empirical Study on Werewolf ๐Ÿ“š

"Keeps LLMs frozen, and relies on retrieval and reflection on past communications and experiences for improvement, which is inspired by the human learning process of "reflection-in-action"." [gal30b+] ๐Ÿค–

๐Ÿ”— arxiv.org/abs/2309.04658v1

#cl #arxiv

Last updated 1 year ago

arXiv Comp. Linguistics๐Ÿ“š · @arxiv_cl
222 followers · 3986 posts · Server creative.ai

๐Ÿ“ Can NLP Models 'Identify', 'Distinguish', and 'Justify' Questions That Don't Have a Definitive Answer? ๐Ÿ“š

"QnotA Dataset is constructed with the aim to test the ability of QA models in identifying, distinguishing and justifying questions without answers (QnotA)." [gal30b+] ๐Ÿค–

๐Ÿ”— arxiv.org/abs/2309.04635v1

#cl #arxiv

Last updated 1 year ago

arXiv Comp. Linguistics๐Ÿ“š · @arxiv_cl
222 followers · 3985 posts · Server creative.ai

๐Ÿ“ When Less Is More: Investigating Data Pruning for Pretraining LLMs at Scale ๐Ÿ“š๐Ÿง 

"Performs a rigorous comparison at scale of the simple data quality estimator of perplexity, as well as more sophisticated and computationally intensive estimates of the Error L2-Norm and memorization." [gal30b+] ๐Ÿค–

๐Ÿ”— arxiv.org/abs/2309.04564v1

#cl #lg #arxiv

Last updated 1 year ago

arXiv Comp. Linguistics๐Ÿ“š · @arxiv_cl
222 followers · 3985 posts · Server creative.ai

๐Ÿ“ MoEController: Instruction-Based Arbitrary Image Manipulation with Mixture-of-Expert Controllers ๐Ÿ”ญ๐Ÿ“š

"Leverages large language models (ChatGPT) and image synthesis models (ControlNet) to generate a large number of image-text pairs that can be used for global and local image manipulation datasets." [gal30b+] ๐Ÿค–

๐Ÿ”— arxiv.org/abs/2309.04372v1

#cv #cl #arxiv

Last updated 1 year ago

arXiv Comp. Linguistics๐Ÿ“š · @arxiv_cl
222 followers · 3983 posts · Server creative.ai

๐Ÿ“ Evaluation and Mitigation of Agnosia in Multimodal Large Language Models ๐Ÿ”ญ๐Ÿ“š

"Proposes EMMA, an evaluation-mitigation framework that automatically creates fine-grained and diverse visual question answering examples to assess the extent of agnosia in Multimodal Pre-trained Language Models (MLLMs) comprehensively." [gal30b+] ๐Ÿค–

๐Ÿ”— arxiv.org/abs/2309.04041v1

#cv #cl #arxiv

Last updated 1 year ago

arXiv Comp. Linguistics๐Ÿ“š · @arxiv_cl
222 followers · 3981 posts · Server creative.ai

๐Ÿ“ Beyond Static Datasets: A Deep Interaction Approach to LLM Evaluation ๐Ÿ“š๐Ÿ‘พ

"Based on the deep interaction between large language models, which can help us evaluate large language models in real-world scenarios such as machine translation and code generation." [gal30b+] ๐Ÿค–

๐Ÿ”— arxiv.org/abs/2309.04369v1

#cl #ai #arxiv

Last updated 1 year ago

arXiv Comp. Linguistics๐Ÿ“š · @arxiv_cl
222 followers · 3980 posts · Server creative.ai

๐Ÿ“ Encoding Multi-Domain Scientific Papers by Ensembling Multiple CLS Tokens ๐Ÿ“š๐Ÿง 

"Proposes Multi2SPE -- it encourages each of multiple CLS tokens to learn diverse ways of aggregating token embeddings, then sums them up together to create a single vector representation." [gal30b+] ๐Ÿค–

๐Ÿ”— arxiv.org/abs/2309.04333v1

#cl #DL #lg #arxiv

Last updated 1 year ago

arXiv Comp. Linguistics๐Ÿ“š · @arxiv_cl
222 followers · 3979 posts · Server creative.ai

๐Ÿ“ Fuzzy Fingerprinting Transformer Language-Models for Emotion Recognition in Conversations ๐Ÿ“š๐Ÿ‘พ

"We feed the utterances and their previous conversational turns to a pre-trained RoBERTa, obtaining contextual embedding utterance representations, that are then supplied to an adapted Fuzzy Fingerprint classification module." [gal30b+] ๐Ÿค–

๐Ÿ”— arxiv.org/abs/2309.04292v1

#cl #ai #arxiv

Last updated 1 year ago

arXiv Comp. Linguistics๐Ÿ“š · @arxiv_cl
222 followers · 3977 posts · Server creative.ai

๐Ÿ“ From Sparse to Dense: GPT-4 Summarization with Chain of Density Prompting ๐Ÿ“š

"Fine-tunes GPT-4 on CNN DailyMail and generate entity-centric summaries by iteratively incorporating missing salient entities without increasing the length of the summary, using a chain of density prompt." [gal30b+] ๐Ÿค–

๐Ÿ”— arxiv.org/abs/2309.04269v1

#cl #arxiv

Last updated 1 year ago

arXiv Comp. Linguistics๐Ÿ“š · @arxiv_cl
222 followers · 3973 posts · Server creative.ai

๐Ÿ“ GLS-CSC: A Simple but Effective Strategy to Mitigate Chinese STM Models' Over-Reliance on Superficial Clue ๐Ÿ“š

"Proposes a novel resampling training strategy called Gradually Learn Samples Containing Superficial Clue (GLS-CSC) to mitigate STM models' over-reliance on superficial clues." [gal30b+] ๐Ÿค–

๐Ÿ”— arxiv.org/abs/2309.04162v1

#cl #arxiv

Last updated 1 year ago

arXiv Comp. Linguistics๐Ÿ“š · @arxiv_cl
222 followers · 3970 posts · Server creative.ai

๐Ÿ“ RST-style Discourse Parsing Guided by Document-Level Content Structures ๐Ÿ“š

"The proposed pipeline for RST-DP incorporates structure-aware news content sentence representations derived from the task of News Discourse Profiling via only a few additional layers in the neural network model architecture." [gal30b+] ๐Ÿค–

๐Ÿ”— arxiv.org/abs/2309.04141v1

#cl #arxiv

Last updated 1 year ago

arXiv Comp. Linguistics๐Ÿ“š · @arxiv_cl
222 followers · 3970 posts · Server creative.ai

๐Ÿ“ Unsupervised Multi-Document Summarization with Holistic Inference ๐Ÿ“š

"Incorporates the holistic beam search inference method associated with the holistic measurements, named Subset Representative Index (SRI), which balances the importance and diversity of a subset of sentences from the source documents and can be calculated in unsupervised and adaptive manners." [gal30b+] ๐Ÿค–

๐Ÿ”— arxiv.org/abs/2309.04087v1

#cl #arxiv

Last updated 1 year ago

arXiv Comp. Linguistics๐Ÿ“š · @arxiv_cl
222 followers · 3970 posts · Server creative.ai

๐Ÿ“ ConDA: Contrastive Domain Adaptation for AI-generated Text Detection ๐Ÿ“š๐Ÿ‘พ๐Ÿง 

"Develops a contrastive domain adaptation framework, called ConDA, which learns domain-invariant feature representations via a contrastive loss in conjunction with standard domain adaptation techniques such as DANN and CDAN." [gal30b+] ๐Ÿค–

โš™๏ธ github.com/AmritaBh/ConDA-gen-
๐Ÿ”— arxiv.org/abs/2309.03992v1

#cl #ai #lg #arxiv

Last updated 1 year ago

arXiv Sound & Audio๐Ÿ”Š · @arxiv_sd
39 followers · 620 posts · Server creative.ai

๐Ÿ“ Multiple Representation Transfer From Large Language Models to End-to-End ASR Systems ๐Ÿ“š๐Ÿ”Š

"Transferring multiple representations of large language models improves the end-to-end ASR performance by up to 15% relative CER compared to transferring only a single representation." [gal30b+] ๐Ÿค–

๐Ÿ”— arxiv.org/abs/2309.04031v1

#cl #sd #arxiv

Last updated 1 year ago

arXiv Comp. Linguistics๐Ÿ“š · @arxiv_cl
221 followers · 3962 posts · Server creative.ai

๐Ÿ“ A Function Interpretation Benchmark for Evaluating Interpretability Methods ๐Ÿ“š๐Ÿ‘พ๐Ÿง 

"We procedurally construct a suite of functions resembling components of trained neural networks, and use them to evaluate interpretability methods that use language models (LMs) to propose descriptions in language or code." [gal30b+] ๐Ÿค–

โš™๏ธ github.com/multimodal-interpre
๐Ÿ”— arxiv.org/abs/2309.03886v1

#cl #ai #lg #arxiv

Last updated 1 year ago

arXiv Sound & Audio๐Ÿ”Š · @arxiv_sd
39 followers · 619 posts · Server creative.ai

๐Ÿ“ Implicit Design Choices and Their Impact on Emotion Recognition Model Development and Evaluation ๐Ÿง ๐Ÿ“š๐Ÿ”Š

"Emotion recognition from visual and audio data is accomplished by using a convolutional neural network (CNN) and a recurrent neural network (RNN), respectively, to encode the two modalities." [gal30b+] ๐Ÿค–

๐Ÿ”— arxiv.org/abs/2309.03238v1

#lg #cl #sd #arxiv

Last updated 1 year ago

arXiv Sound & Audio๐Ÿ”Š · @arxiv_sd
39 followers · 618 posts · Server creative.ai

๐Ÿ“ Zero-Shot Audio Captioning via Audibility Guidance ๐Ÿ”Š๐Ÿ“š

"A caption is generated by combining a large pre-trained language model, such as GPT-2, with a multimodal matching model, which scores how well a text matches the input audio and a text classifier which provides the guidance for audibility." [gal30b+] ๐Ÿค–

๐Ÿ”— arxiv.org/abs/2309.03884v1

#sd #cl #arxiv

Last updated 1 year ago

๐Ÿ“ Introducing "Forecast Utterance" for Conversational Data Science ๐Ÿ“š๐Ÿ‘‹

"Introduces a new concept called a Forecast Utterance and then focus on automatically interpreting users' prediction goals from these Utterances by casting it as a slot-filling problem." [gal30b+] ๐Ÿค–

๐Ÿ”— arxiv.org/abs/2309.03877v1

#cl #hc #arxiv

Last updated 1 year ago

arXiv Comp. Linguistics๐Ÿ“š · @arxiv_cl
220 followers · 3954 posts · Server creative.ai

๐Ÿ“ The Daunting Dilemma with Sentence Encoders: Success on Standard Benchmarks, Failure in Capturing Basic Semantic Properties ๐Ÿ“š

"Presents a study to explore the performance of popular sentence encoders on downstream tasks and their capability to capture basic semantic properties such as paraphrasing, synonym replacement, antonym replacement, and jumbling." [gal30b+] ๐Ÿค–

๐Ÿ”— arxiv.org/abs/2309.03747v1

#cl #arxiv

Last updated 1 year ago

arXiv Comp. Linguistics๐Ÿ“š · @arxiv_cl
220 followers · 3953 posts · Server creative.ai

๐Ÿ“ Word Segmentation Granularity in Korean ๐Ÿ“š

"The process of splitting a sentence into words for further processing, for instance, part-of-speech tagging, chunking, dependency parsing, etc." [gal30b+] ๐Ÿค–

๐Ÿ”— arxiv.org/abs/2309.03713v1

#cl #arxiv

Last updated 1 year ago