TechcraftingAI NLP

TechcraftingAI NLP

By Brad EdwardsTechnology
Download on the App Store

All TechcraftingAI NLP episodes

  • Ep. 256 - Part 1 - June 6, 2024

    ArXiv NLP research for Thursday, June 06, 2024.


    00:20: Efficient Knowledge Infusion via KG-LLM Alignment

    01:25: NAP^2: A Benchmark for Naturalness and Privacy-Preserving Text Rewriting by Learning from Human

    02:34: Character-Level Chinese Dependency Parsing via Modeling Latent Intra-Word Structure

    03:30: XL-HeadTags: Leveraging Multimodal Retrieval Augmentation for the Multilingual Generation of News Headlines and Tags

    04:59: End-to-End Trainable Soft Retriever for Low-resource Relation Extraction

    06:07: Light-PEFT: Lightening Parameter-Efficient Fine-Tuning via Early Pruning

    07:37: Improving Zero-Shot Chinese-English Code-Switching ASR with kNN-CTC and Gated Monolingual Datastores

    08:52: ReST-MCTS*: LLM Self-Training via Process Reward Guided Tree Search

    10:29: Chaos with Keywords: Exposing Large Language Models Sycophancy to Misleading Keywords and Evaluating Defense Strategies

    11:39: Lean Workbook: A large-scale Lean problem set formalized from natural language math problems

    12:56: Speculative Decoding via Early-exiting for Faster LLM Inference with Thompson Sampling Control Mechanism

    14:18: Performance of large language models in numerical vs. semantic medical knowledge: Benchmarking on evidence-based Q&As

    16:24: Recovering document annotations for sentence-level bitext

    17:40: BLSP-Emo: Towards Empathetic Large Speech-Language Models

    19:01: Decoder-only Streaming Transformer for Simultaneous Translation

    20:28: Evaluating the IWSLT2023 Speech Translation Tasks: Human Annotations, Automatic Metrics, and Segmentation

    21:53: Spontaneous Speech-Based Suicide Risk Detection Using Whisper and Large Language Models

    23:06: How Good is Zero-Shot MT Evaluation for Low Resource Indian Languages?

    24:13: HeSum: a Novel Dataset for Abstractive Text Summarization in Hebrew

    25:19: ArMeme: Propagandistic Content in Arabic Memes

    26:26: Culturally Aware and Adapted NLP: A Taxonomy and a Survey of the State of the Art

    27:11: UltraMedical: Building Specialized Generalists in Biomedicine

    28:43: Tox-BART: Leveraging Toxicity Attributes for Explanation Generation of Implicit Hate Speech

    30:02: A + B: A General Generator-Reader Framework for Optimizing LLMs to Unleash Synergy Potential

    31:29: On The Persona-based Summarization of Domain-Specific Documents

    33:14: Assessing LLMs for Zero-shot Abstractive Summarization Through the Lens of Relevance Paraphrasing

    34:28: American Sign Language Handshapes Reflect Pressures for Communicative Efficiency

    36 min
  • Ep. 255 - June 5, 2024

    ArXiv NLP research for Wednesday, June 05, 2024.


    00:19: Improving In-Context Learning with Prediction Feedback for Sentiment Analysis

    01:24: MultifacetEval: Multifaceted Evaluation to Probe LLMs in Mastering Medical Knowledge

    03:01: Text Injection for Neural Contextual Biasing

    04:16: 4D ASR: Joint Beam Search Integrating CTC, Attention, Transducer, and Mask Predict Decoders

    06:03: Adversarial Moment-Matching Distillation of Large Language Models

    07:05: Docs2KG: Unified Knowledge Graph Construction from Heterogeneous Documents Assisted by Large Language Models

    08:48: Readability-guided Idiom-aware Sentence Simplification (RISS) for Chinese

    09:56: Evaluation of data inconsistency for multi-modal sentiment analysis

    10:55: BadAgent: Inserting and Activating Backdoor Attacks in LLM Agents

    12:11: Unveiling Selection Biases: Exploring Order and Token Sensitivity in Large Language Models

    13:16: From Tarzan to Tolkien: Controlling the Language Proficiency Level of LLMs for Content Generation

    14:20: StreamSpeech: Simultaneous Speech-to-Speech Translation with Multi-task Learning

    15:42: RadBARTsum: Domain Specific Adaption of Denoising Sequence-to-Sequence Models for Abstractive Radiology Report Summarization

    17:00: Towards Detecting LLMs Hallucination via Markov Chain-based Multi-agent Debate Framework

    18:14: Cryptocurrency Frauds for Dummies: How ChatGPT introduces us to fraud?

    19:48: FragRel: Exploiting Fragment-level Relations in the External Memory of Large Language Models

    20:59: Space Decomposition for Sentence Embedding

    22:00: Towards Real-world Scenario: Imbalanced New Intent Discovery

    23:40: Which Side Are You On? A Multi-task Dataset for End-to-End Argument Summarisation and Evaluation

    25:20: CSS: Contrastive Semantic Similarity for Uncertainty Quantification of LLMs

    27:03: StatBot.Swiss: Bilingual Open Data Exploration in Natural Language

    28:10: Missci: Reconstructing Fallacies in Misrepresented Science

    29:43: ChatLang-8: An LLM-Based Synthetic Data Generation Framework for Grammatical Error Correction

    30:47: Linking Named Entities in Diderot's \textit{Encyclop\'edie} to Wikidata

    32:06: Error-preserving Automatic Speech Recognition of Young English Learners' Language

    33:37: Document-level Claim Extraction and Decontextualisation for Fact-Checking

    34:45: The Challenges of Evaluating LLM Applications: An Analysis of Automated, Human, and LLM-Based Approaches

    36:09: LLM-based Rewriting of Inappropriate Argumentation using Reinforcement Learning from Machine Feedback

    37:39: IrokoBench: A New Benchmark for African Languages in the Age of Large Language Models

    39:46: Automating Turkish Educational Quiz Generation Using Large Language Models

    41:34: Cycles of Thought: Measuring LLM Confidence through Stable Explanations

    42:57: Are language models rational? The case of coherence norms and belief revision

    43:58: What is the Best Way for ChatGPT to Translate Poetry?

    45:20: Using Synchronic Definitions and Semantic Relations to Classify Semantic Change Types

    46:14: MODABS: Multi-Objective Learning for Dynamic Aspect-Based Summarization

    47:09: BIPED: Pedagogically Informed Tutoring System for ESL Education

    48:24: Analyzing LLM Behavior in Dialogue Summarization: Unveiling Circumstantial Hallucination Trends

    50:00: Wings: Learning Multimodal LLMs without Text-only Forgetting

    52 min
  • Ep. 254 - Part 2 - June 4, 2024

    ArXiv NLP research for Tuesday, June 04, 2024.


    00:20: Description Boosting for Zero-Shot Entity and Relation Classification

    01:44: Modeling Emotional Trajectories in Written Stories Utilizing Transformers and Weakly-Supervised Learning

    03:09: Enhancing Retrieval-Augmented LMs with a Two-stage Consistency Learning Compressor

    04:30: Prompting Large Language Models with Human Error Markings for Self-Correcting Machine Translation

    05:41: mCoT: Multilingual Instruction Tuning for Reasoning Consistency in Language Models

    06:53: Technical Language Processing for Telecommunications Specifications

    08:09: On Affine Homotopy between Language Encoders

    09:25: Translation Deserves Better: Analyzing Translation Artifacts in Cross-lingual Visual Question Answering

    10:32: Probing the Category of Verbal Aspect in Transformer Language Models

    11:58: Linguistic Fingerprint in Transformer Models: How Language Variation Influences Parameter Selection in Irony Detection

    13:03: LlamaCare: A Large Medical Language Model for Enhancing Healthcare Knowledge Sharing

    14:33: Retaining Key Information under High Compression Ratios: Query-Guided Compressor for LLMs

    15:51: On the Intrinsic Self-Correction Capability of LLMs: Uncertainty and Latent Concept

    17:30: Multiple Choice Questions and Large Languages Models: A Case Study with Fictional Medical Data

    19:08: The Scandinavian Embedding Benchmarks: Comprehensive Assessment of Multilingual and Monolingual Text Embedding

    20:07: Representations as Language: An Information-Theoretic Framework for Interpretability

    21:32: Analyzing Temporal Complex Events with Large Language Models? A Benchmark towards Temporal, Long Context Understanding

    22:46: Hiding Text in Large Language Models: Introducing Unconditional Token Forcing Confusion

    24:21: Language-Universal Speech Attributes Modeling for Zero-Shot Multilingual Spoken Keyword Recognition

    25:37: Deterministic Reversible Data Augmentation for Neural Machine Translation

    26:39: CheckEmbed: Effective Verification of LLM Solutions to Open-Ended Tasks

    28:14: Scalable MatMul-free Language Modeling

    30:03: SpecExec: Massively Parallel Speculative Decoding for Interactive LLM Inference on Consumer Devices

    31:37: Mitigate Position Bias in Large Language Models via Scaling a Single Dimension

    33:10: TopViewRS: Vision-Language Models as Top-View Spatial Reasoners

    35 min
  • Ep. 254 - Part 1 - June 4, 2024

    ArXiv NLP research for Tuesday, June 04, 2024.


    00:20: Conditional Language Learning with Context

    01:13: Zyda: A 1.3T Dataset for Open Language Modeling

    02:32: RKLD: Reverse KL-Divergence-based Knowledge Distillation for Unlearning Personal Information in Large Language Models

    03:50: Personalized Topic Selection Model for Topic-Grounded Dialogue

    05:20: Position Debiasing Fine-Tuning for Causal Perception in Long-Term Dialogue

    06:58: Phonetic Enhanced Language Modeling for Text-to-Speech Synthesis

    08:03: Why Would You Suggest That? Human Trust in Language Model Responses

    09:10: Multimodal Reasoning with Multimodal Knowledge Graph

    10:30: QROA: A Black-Box Query-Response Optimization Attack on LLMs

    11:55: Analyzing Social Biases in Japanese Large Language Models

    12:52: I've got the "Answer"! Interpretation of LLMs Hidden States in Question Answering

    13:47: PyramidKV: Dynamic KV Cache Compression based on Pyramidal Information Funneling

    15:16: Assessing the Performance of Chinese Open Source Large Language Models in Information Extraction Tasks

    16:38: LongSSM: On the Length Extension of State-space Models in Language Modelling

    17:30: Exploring Mathematical Extrapolation of Large Language Models with Synthetic Data

    18:40: MARS: Benchmarking the Metaphysical Reasoning Abilities of Language Models with a Multi-task Evaluation Dataset

    20:19: UniOQA: A Unified Framework for Knowledge Graph Question Answering with Large Language Models

    22:03: Diver: Large Language Model Decoding with Span-Level Mutual Information Verification

    23:12: SimulTron: On-Device Simultaneous Speech to Speech Translation

    24:28: The current status of large language models in summarizing radiology report impressions

    26:10: Reinforcement Tuning for Detecting Stances and Debunking Rumors Jointly with Large Language Models

    27:17: Synergetic Event Understanding: A Collaborative Approach to Cross-Document Event Coreference Resolution with Large Language Models

    28:46: A multilingual dataset for offensive language and hate speech detection for hausa, yoruba and igbo languages

    29:40: FedMKT: Federated Mutual Knowledge Transfer for Large and Small Language Models

    31:17: Self-Modifying State Modeling for Simultaneous Machine Translation

    34 min
  • Ep. 253 - June 3, 2024

    ArXiv NLP research for Monday, June 03, 2024.


    00:19: Luna: An Evaluation Foundation Model to Catch Language Model Hallucinations with High Accuracy and Low Cost

    01:38: Generative Pre-trained Speech Language Model with Efficient Hierarchical Transformer

    03:06: Selectively Answering Visual Questions

    04:11: Take its Essence, Discard its Dross! Debiasing for Toxic Language Detection via Counterfactual Causal Effect

    05:36: Predicting Drug-Gene Relations via Analogy Tasks with Word Embeddings

    06:51: SemCoder: Training Code Language Models with Comprehensive Semantics

    08:39: Mobile-Agent-v2: Mobile Device Operation Assistant with Effective Navigation via Multi-Agent Collaboration

    10:26: Combining Qualitative and Computational Approaches for Literary Analysis of Finnish Novels

    11:45: Strengthened Symbol Binding Makes Large Language Models Reliable Multiple-Choice Selectors

    13:26: Decompose, Enrich, and Extract! Schema-aware Event Extraction using LLMs

    14:34: MACT: Model-Agnostic Cross-Lingual Training for Discourse Representation Structure Parsing

    15:48: Guiding ChatGPT to Generate Salient Domain Summaries

    17:51: Synergizing Unsupervised and Supervised Learning: A Hybrid Approach for Accurate Natural Language Task Modeling

    19:30: TCMBench: A Comprehensive Benchmark for Evaluating Large Language Models in Traditional Chinese Medicine

    21:38: Explore then Determine: A GNN-LLM Synergy Framework for Reasoning over Knowledge Graph

    22:51: Two Tales of Persona in LLMs: A Survey of Role-Playing and Personalization

    24:08: Are AI-Generated Text Detectors Robust to Adversarial Perturbations?

    25:42: Automatic Essay Multi-dimensional Scoring with Fine-tuning and Multiple Regression

    26:35: Improving Pseudo Labels with Global-Local Denoising Framework for Cross-lingual Named Entity Recognition

    28:01: Demonstration Augmentation for Zero-shot In-context Learning

    29:31: EffiQA: Efficient Question-Answering with Strategic Multi-Model Collaboration on Knowledge Graphs

    31:05: Towards Scalable Automated Alignment of LLMs: A Survey

    32:19: EduNLP: Towards a Unified and Modularized Library for Educational Resources

    33:44: Focus on the Core: Efficient Attention via Pruned Token Compression for Document Classification

    35:07: Improved Few-Shot Jailbreaking Can Circumvent Aligned Language Models and Their Defenses

    36:36: When Can LLMs Actually Correct Their Own Mistakes? A Critical Survey of Self-Correction of LLMs

    37:58: CodeR: Issue Resolving with Multi-Agent and Task Graphs

    38:54: Unsupervised Distractor Generation via Large Language Model Distilling and Counterfactual Contrastive Decoding

    40:10: FactGenius: Combining Zero-Shot Prompting and Fuzzy Relation Mining to Improve Fact Verification with Knowledge Graphs

    41:27: Probing Language Models for Pre-training Data Detection

    42:45: R2C2-Coder: Enhancing and Benchmarking Real-world Repository-level Code Completion Abilities of Code Large Language Models

    44:32: Privacy in LLM-based Recommendation: Recent Advances and Future Directions

    45:23: Linguistic Analysis, Description, and Typological Exploration with Categorial Grammar (TheBench Guide)

    46:52: D-CPT Law: Domain-specific Continual Pre-Training Scaling Law for Large Language Models

    48:52: Do Large Language Models Perform the Way People Expect? Measuring the Human Generalization Function

    50:07: Sparsity-Accelerated Training for Large Language Models

    51:36: Superhuman performance in urology board questions by an explainable large language model enabled for context integration of the European Association of Urology guidelines: the UroBot study

    53:34: Editing the Mind of Giants: An In-Depth Exploration of Pitfalls of Knowledge Editing in Large Language Models

    54:42: LexMatcher: Dictionary-centric Data Collection for LLM-based Machine Translation

    55:55: Enabling ASR for Low-Resource Languages: A Comprehensive Dataset Creation Approach

    57:10: Understanding Token Probability Encoding in Output Embeddings



    1 hr 11 min
  • Ep. 252 - June 2, 2024

    ArXiv NLP research for Sunday, June 02, 2024.


    00:19: Prompt Framework for Role-playing: Generation and Evaluation

    01:05: Transforming Computer Security and Public Trust Through the Exploration of Fine-Tuning Large Language Models

    02:18: Enhancing Zero-shot Text-to-Speech Synthesis with Human Feedback

    03:54: Presence or Absence: Are Unknown Word Usages in Dictionaries?

    05:09: Topic Modeling for Short Texts with Large Language Models

    06:09: How well do distributed representations convey contextual lexical semantics: a Thesis Proposal

    07:05: Evaluating Mathematical Reasoning of Large Language Models: A Focus on Error Identification and Correction

    08:27: Automatic Instruction Evolving for Large Language Models

    09:25: Applying Intrinsic Debiasing on Downstream Tasks: Challenges and Considerations for Machine Translation

    10:26: Developing an efficient corpus using Ensemble Data cleaning approach

    11:51: BoNBoN Alignment for Large Language Models and the Sweetness of Best-of-n Sampling

    13:15: FOCUS: Forging Originality through Contrastive Use in Self-Plagiarism for Language Models

    14:51: The Power of Summary-Source Alignments

    16:11: Formality Style Transfer in Persian

    17:39: Show, Don't Tell: Aligning Language Models with Demonstrated Feedback

    19:08: YODAS: Youtube-Oriented Dataset for Audio and Speech

    20:13: MEDIQ: Question-Asking LLMs for Adaptive and Reliable Medical Reasoning

    22:15: A Survey of Useful LLM Evaluation

    23:31: Unveil the Duality of Retrieval-Augmented Generation: Theoretical Analysis and Practical Solution

    25:07: Annotation Guidelines-Based Knowledge Augmentation: Towards Enhancing Large Language Models for Educational Text Classification

    27:18: Using RL to Identify Divisive Perspectives Improves LLMs Abilities to Identify Communities on Social Media

    29 min
  • Ep. 251 - June 1, 2024

    ArXiv NLP research for Saturday, June 01, 2024.


    00:19: Multi-Dimensional Optimization for Text Summarization via Reinforcement Learning

    01:41: CASE: Curricular Data Pre-training for Building Generative and Discriminative Assistive Psychology Expert Models

    03:25: Beyond Metrics: Evaluating LLMs' Effectiveness in Culturally Nuanced, Low-Resource Real-World Scenarios

    05:03: RoBERTa-BiLSTM: A Context-Aware Hybrid Model for Sentiment Analysis

    07:09: The Best of Both Worlds: Toward an Honest and Helpful Large Language Model

    09:02: Gender Bias Detection in Court Decisions: A Brazilian Case Study

    10:41: Prompt Chaining or Stepwise Prompt? Refinement in Text Summarization

    11:54: A Survey on Large Language Models for Code Generation

    13:43: Guiding and Diversifying LLM-Based Story Generation via Answer Set Programming

    14:46: SPAGHETTI: Open-Domain Question Answering from Heterogeneous Data Sources with Retrieval and Semantic Parsing

    15:43: LongSkywork: A Training Recipe for Efficiently Extending Context Length in Large Language Models

    17:24: LLMs Could Autonomously Learn Without External Supervision

    19 min
  • Ep. 250 - May 31, 2024

    ArXiv NLP research summaries for May 31, 2024.


    00:20 FineRadScore: A Radiology Report Line-by-Line Evaluation Technique Generating Corrections with Severity Scores

    01:37 Leveraging Large Language Models for Entity Matching

    02:27 Reward-based Input Construction for Cross-document Relation Extraction

    03:40 Passage-specific Prompt Tuning for Passage Reranking in Question Answering with Large Language Models

    05:04 DORY: Deliberative Prompt Recovery for LLM

    06:18 Unveiling the Lexical Sensitivity of LLMs: Combinatorial Optimization for Prompt Enhancement

    07:35 It is Simple Sometimes: A Study On Improving Aspect-Based Sentiment Analysis Performance

    08:59 FinGen: A Dataset for Argument Generation in Finance

    09:42 Improving code-mixed hate detection by native sample mixing: A case study for Hindi-English code-mixed scenario

    11:26 Multilingual Text Style Transfer: Datasets & Models for Indian Languages

    13:01 An iterated learning model of language change that mixes supervised and unsupervised learning

    14:01 Self-Augmented Preference Optimization: Off-Policy Paradigms for Language Model Alignment

    15:29 That's Optional: A Contemporary Exploration of "that" Omission in English Subordinate Clauses

    16:18 Don't Buy it! Reassessing the Ad Understanding Abilities of Contrastive Multimodal Models

    17:20 Improving Reward Models with Synthetic Critiques

    18:29 Towards Spoken Language Understanding via Multi-level Multi-grained Contrastive Learning

    19:49 clembench-2024: A Challenging, Dynamic, Complementary, Multilingual Benchmark and Underlying Flexible Framework for LLMs as Multi-Action Agents

    21:05 A comparison of correspondence analysis with PMI-based word embedding methods

    22:05 Large Language Models: A New Approach for Privacy Policy Analysis at Scale

    23:36 Preemptive Answer "Attacks" on Chain-of-Thought Reasoning

    24:22 Learning to Estimate System Specifications in Linear Temporal Logic using Transformers and Mamba

    25:48 OR-Bench: An Over-Refusal Benchmark for Large Language Models

    27:20 Superlatives in Context: Explicit and Implicit Domain Restrictions for Superlative Frames

    28:41 SaySelf: Teaching LLMs to Express Confidence with Self-Reflective Rationales

    30:33 Towards a Fluid computer

    31:33 You Only Scan Once: Efficient Multi-dimension Sequential Modeling with LightNet

    33:01 LACIE: Listener-Aware Finetuning for Confidence Calibration in Large Language Models

    35:02 Direct Alignment of Language Models via Quality-Aware Self-Refinement

    36:19 Code Pretraining Improves Entity Tracking Abilities of Language Models

    38 min