The implementation of our paper accepted by ACL 2023: CASE: Aligning Coarse-to-Fine Cognition and Affection for Empathetic Response Generation

An AI-chat bot transforming counseling with personalized support and expert assistance.

Python program to make LLM agents play characters and talk to each other.

Aim 💫 — An easy-to-use & supercharged open-source experiment tracker.

LLM Roleplay: Simulating Human-Chatbot Interaction

[EMNLP 2024] ”ESC-Eval: Evaluating Emotion Support Conversations in Large Language Models“

SCOPE ICLR 2025

ConvBench: A Multi-Turn Conversation Evaluation Benchmark with Hierarchical Ablation Capability for Large Vision-Language Models

This repository contains code for the EMNLP 2025 Findings paper [Chain of Strategy Optimization Makes Large Language Models Better Emotional Supporter](https://arxiv.org/abs/2503.05362)

The implementation of <Factual Consistency Evaluation for Text Summarization via Counterfactual Estimation> in PyTorch.

"MiniRAG: Making RAG Simpler with Small and Open-Sourced Language Models"

Code that accompanies the public release of the paper Lost in Conversation (https://arxiv.org/abs/2505.06120)

TrustAgent: Towards Safe and Trustworthy LLM-based Agents

Toolkit for evaluating the trustworthiness of generative foundation models.

[ICLR'25] DataGen: Unified Synthetic Dataset Generation via Large Language Models

[ICML 2024] TrustLLM: Trustworthiness in Large Language Models

Codes and datasets for the paper Measuring and Enhancing Trustworthiness of LLMs in RAG through Grounded Attributions and Learning to Refuse

The LLM Evaluation Framework

Python client library for Cleanlab Trustworthy Language Model

Crowd-sourced lists of urls to help Common Crawl crawl under-resourced languages. See https://github.com/commoncrawl/web-languages-code/ for the code

implementation for retrieval-based dialog system using pytorch pretrained BERT.

Source Code for DialogBERT: Discourse-Aware Response Generation via Learning to Recover and Rank Utterances (https://arxiv.org/pdf/2012.01775.pdf)

DialogueSentenceBERT: SentenceBERT for More Representative Utterance Embedding via Pre-training on Dialogue Corpus

Pre-Trained Models for ToD-BERT

Code and data for "Reasoning Like a Doctor: Improving Medical Dialogue Systems via Diagnostic Reasoning Process Alignment" (ACL 2024 Findings)

"RAG-Anything: All-in-One RAG Framework"

GraphRAG-Bench, the official repo of comprehensive benchmark and dataset for evaluating GraphRAG models.