[EMNLP 2025 main] C3 Benchmark: A Bilingual Benchmark for Spoken Dialogue Models Exploring Challenges in Complex Conversations

A handy tool to help design and parse your own language

Code and Assets for "Benchmarking and Improving Text-to-SQL Generation Under Ambiguity" (EMNLP 2023)

Drainage: A Unifying Framework for Addressing Class Uncertainty

🐶 Data for AmbigNLG: Addressing Task Ambiguity in Instruction for NLG (Niwa and Iso; EMNLP 2024)

Pytorch implementation of "Aligning Language Models to Explicitly Handle Ambiguity" (EMNLP 2024)

Code for the paper Sentence Ambiguity, Grammaticality and Complexity Probes

Resolving Ambiguities in Text-to-SQL Systems . We propose the RAQS-SQL framework to Resolve Ambiguities in QuestionS for Text-to-SQL system. To handle schema-level ambiguity, we use a model that aligns the query intent directly with relevant database columns.

Utilities for lark's ambiguous trees

Official code for "An Extensible Benchmark for Value Ambiguity Resolution in Text-to-SQL"

ez-convphase - for easily phasing most fasta alignments

Do aligned LLMs recognize a sentence's second reading but not produce it? A reproducible pilot measuring the recognition-vs-generation gap — and whether temperature or a random seed-prefix recovers the flattened reading (Qwen3.5 · gpt-5.5 forced-choice scoring).

ClariGen is a multi-model pipeline designed to detect ambiguous user queries and generate precise clarifying questions. It leverages a tiered LLM approach to balance speed and intelligence.