Fangcong Yin

Randomized-YaRN. Python

VRRL. Official codebase for VRRL: Visual Grounded Self-Reflection for VLMs via RL

Dr-Claude-Code. Python

Gelato. 🍨 Gelato — From Data Curation to Reinforcement Learning: Building a Strong Grounding Model for Computer-Use Agents

SkillNet. Create, Evaluate, and Connect AI Skills

skillsbench. SkillsBench evaluates how well skills work and how effective agents are at using them.

DySCO. DySCO: Dynamic Attention-Scaling Decoding for Long-Context LMs

GroundCUA. GroundCUA

STAT. Skill-Targeted Adaptive Training

composable_cot. Python

openrlhf-pretrain. Code for "Echo Chamber: RL Post-training Amplifies Behaviors Learned in Pretraining"

darling. Official Implementation of the paper "Jointly Reinforcing Diversity and Quality in Language Model Generations"

EasySteer. A Unified Framework for High-Performance and Extensible LLM Steering

open-thoughts. Fully open data curation for reasoning models

VTool-R1. [ICLR 2026] "VTool-R1: VLMs Learn to Think with Images via Reinforcement Learning on Multimodal Tool Use"

OrionBench. Jupyter Notebook

octotools. OctoTools: An agentic framework with extensible tools for complex reasoning

ChartMuseum. [NeurIPS 2025] ChartMuseum: Testing Visual Reasoning Capabilities of Large Vision-Language Models

CharXiv. [NeurIPS 2024] CharXiv: Charting Gaps in Realistic Chart Understanding in Multimodal LLMs

sam2. The repository provides code for running inference with the Meta Segment Anything Model 2 (SAM 2), links for downloading the trained model checkpoints, and example notebooks that show how to use the model.

ReFocus_Code. Codes for ReFocus: Visual Editing as a Chain of Thought for Structured Image Understanding [ICML 2025]]

OpenThinkIMG. OpenThinkIMG is an end-to-end open-source framework that empowers Large Vision-Language Models to think with images.

Awesome_Think_With_Images. Resources and paper list for "Thinking with Images for LVLMs". This repository accompanies our survey on how LVLMs can leverage visual information for complex reasoning, planning, and generation.

QRHead. QRHead: Query-Focused Retrieval Heads Improve Long-Context Reasoning and Re-ranking

knowledge-graph-of-thoughts. Official Implementation of "Affordable AI Assistants with Knowledge Graph of Thoughts"

OpenRLHF. An Easy-to-use, Scalable and High-performance Agentic RL Framework based on Ray (PPO & DAPO & REINFORCE++ & VLM & TIS & vLLM & Ray & Async RL)

Long-to-Short-via-Model-Merging. Model merging is a highly efficient approach for long-to-short reasoning.

STAR. Repo for "STAR: Spectral Truncation and Rescale for Model Merging" [NAACL 2025]

tensor2struct-public. Semantic parsers based on encoder-decoder framework

Localize-and-Stitch. Localize-and-Stitch: Efficient Model Merging via Sparse Task Arithmetic

axbench. Stanford NLP Python library for benchmarking the utility of LLM interpretability methods

LLM-Adapters. Code for our EMNLP 2023 Paper: "LLM-Adapters: An Adapter Family for Parameter-Efficient Fine-Tuning of Large Language Models"

fusion_bench. FusionBench: A Comprehensive Benchmark/Toolkit of Deep Model Fusion

AdaMerging. AdaMerging: Adaptive Model Merging for Multi-Task Learning. ICLR, 2024.

HELMET. The HELMET Benchmark

LongProc. LongProc: Benchmarking Long-Context Language Models on Long Procedural Generation

To-CoT-or-not-to-CoT. Python

lo-fit. LoFiT: Localized Fine-tuning on LLM Representations

LlamaFactory. Unified Efficient Fine-Tuning of 100+ LLMs & VLMs (ACL 2024)

activation-steering. [ICLR 2025] General-purpose activation steering library

synthetic_data_retrieval_heads.

OLMoE. OLMoE: Open Mixture-of-Experts Language Models

lm-evaluation-harness. A framework for few-shot evaluation of language models.

Retrieval_Head. open-source code for paper: Retrieval Head Mechanistically Explains Long-Context Factuality

finetuning. This repository contains the code used for the experiments in the paper "Fine-Tuning Enhances Existing Mechanisms: A Case Study on Entity Tracking".

investigating-alignment. Code and data for the paper: "From Distributional to Overton Pluralism: Investigating Large Language Model Alignment"

lost-in-the-middle. Code and data for "Lost in the Middle: How Language Models Use Long Contexts"

Awesome-Attention-Heads. An awesome repository & A comprehensive survey on interpretability of LLM attention heads.

Minitron. A family of compressed models obtained via pruning and knowledge distillation

SAT-LM. SatLM: SATisfiability-Aided Language Models using Declarative Prompting (NeurIPS 2023)

Logic-LLM. The project page for "LOGIC-LM: Empowering Large Language Models with Symbolic Solvers for Faithful Logical Reasoning"

pal. PaL: Program-Aided Language Models (ICML 2023)

DCR. Python

CodeUpdateArena. Python

pyreft. Stanford NLP Python library for Representation Finetuning (ReFT)

GSM-Plus. GSM-Plus: Data, Code, and Evaluation for Enhancing Robust Mathematical Reasoning in Math Word Problems.

tuned-lens. Tools for understanding how transformer predictions are built layer-by-layer

TOXIGEN. This repo contains the code for generating the ToxiGen dataset, published at ACL 2022.

arxiv-latex-cleaner. arXiv LaTeX Cleaner: Easily clean the LaTeX code of your paper to submit to arXiv

MiniCheck. MiniCheck: Efficient Fact-Checking of LLMs on Grounding Documents [EMNLP 2024]

MuSR. Python

honest_llama. Inference-Time Intervention: Eliciting Truthful Answers from a Language Model

fp-dataset-artifacts. Final project starter code for NLP classes at UT Austin: provides a HuggingFace trainer to enable studying of dataset artifacts.

LIT_auto-gen-contrast-set. Python

baukit. Python