MEGAnno in Action: Human-LLM Collaborative AnnotationMEGAnno combines the power of large language models (LLMs) with human expertise to streamline and enhance the data labeling process with a…Sep 9Sep 9
Leveraging LLMs for Semantic Type Detection in Data LakesWant to exploit more of the data in your data lake? LLMs can help you find data, identify data types, and more. Here’s how…Aug 29Aug 29
Big Trends Shaping NLP & NAACL 2024NAACL showcased major trends: targeted evaluation, reasoning, and fine-tuning/retrieval-augmented generation (RAG). These trends…Aug 16Aug 16
Unlocking the Potential of Transformers for Long-Form Text Matching: A Simple yet Powerful ApproachWe propose a simple yet effective solution using sequence pair classification with Transformer models, demonstrating its superiority over…Aug 15Aug 15
Towards Enterprise Compound AI SystemsWe introduce three projects: (1) developing a suitable architecture for productizing compound AI systems, (2) optimizing agentic workflowsJul 24Jul 24
Order Matters: Assessing LLM Sensitivity in Multiple-Choice TasksLarge language models (LLMs) have demonstrated remarkable capabilities in various NLP tasks. However, previous works have shown these…Jul 17Jul 17
Deep Dive with WiTQA: When Does Retrieval Augmentation Help (or Hurt) Language Models?We built a new question-answering dataset called WiTQA and comprehensively evaluated language models of varying sizes in conjunction…Jul 17Jul 17
Human-LLM Collaborative Annotation Through Effective Verification of LLM LabelsEffectively utilize LLMs as collaborators for data annotation to improve accuracy. We propose a multi-step human-LLM collaborative…May 9May 9
Less Is More for Long Document Summary Evaluation by LLMsOur “Extract then Evaluate” approach, is a significant leap forward in evaluating long document summaries. Enhances efficiency and…Mar 29Mar 29
LLMs as Data Annotators (Part 2) — MEGAnno+: A Human-LLM Collaborative Annotation SystemWe introduce our human-LLM collaborative annotation tool, MEGAnno+, addressing the challenges in LLM annotation by integrating human expertsMar 18Mar 18