debjitpaul/refiner
About The corresponding code from our paper " REFINER: Reasoning Feedback on Intermediate Representations" (EACL 2024). Do not hesitate to open an issue if you run into any trouble!
This project helps improve the accuracy of language models on complex reasoning tasks by providing structured feedback on intermediate steps. You input a problem, such as a math word problem or a moral dilemma, and it outputs the refined reasoning steps and a more accurate final answer. This is useful for researchers and practitioners working on natural language understanding and generation who need to enhance the reliability of AI systems for tasks requiring logical thought.
Use this if you need to build or evaluate language models that can accurately perform multi-step reasoning, such as solving math word problems or making moral judgments, by improving their intermediate thought processes.
Not ideal if you are looking for a plug-and-play solution for simple text generation or classification tasks that do not require complex, explainable reasoning steps.
Stars
74
Forks
11
Language
Python
License
Apache-2.0
Category
Last pushed
Jan 27, 2026
Commits (30d)
0
Get this data via API
curl "https://pt-edge.onrender.com/api/v1/quality/nlp/debjitpaul/refiner"
Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.
Related tools
THUDM/P-tuning
A novel method to tune language models. Codes and datasets for paper ``GPT understands, too''.
ZixuanKe/PyContinual
PyContinual (An Easy and Extendible Framework for Continual Learning)
arazd/ProgressivePrompts
Progressive Prompts: Continual Learning for Language Models
Nithin-Holla/MetaLifelongLanguage
Repository containing code for the paper "Meta-Learning with Sparse Experience Replay for...
zjunlp/ContinueMKGC
[IJCAI 2024] Continual Multimodal Knowledge Graph Construction