A bibliography and survey of the papers surrounding o1

[ICLR 2025 Spotlight] Official Implementation for ToST (Token Statistics Transformer)

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

Repo for the research paper "SecAlign: Defending Against Prompt Injection with Preference Optimization"

Code implementation of synthetic continued pretraining

Forecasting with LLMs

xLSTM as Generic Vision Backbone

Pytorch implementation of Transfusion, "Predict the Next Token and Diffuse Images with One Multi-Modal Model", from MetaAI

The official implementation of DiM: Diffusion Mamba for Efficient High-Resolution Image Synthesis

Efficient Triton Kernels for LLM Training

Official repo for consistency models.

20+ high-performance LLMs with recipes to pretrain, finetune and deploy at scale.

[ICML 2024] CLLMs: Consistency Large Language Models

Official PyTorch Implementation of the Longhorn Deep State Space Model

Official Implementation of Rectified Flow (ICLR2023 Spotlight)

Codebase for the ICML 2024 paper "Differentially Private Representation Learning via Image Captioning"

Universal and Transferable Attacks on Aligned Language Models

Code for the paper "Interpreting and Improving Diffusion Models from an Optimization Perspective", appearing in ICML 2024

The official implementation for Pseudo Numerical Methods for Diffusion Models on Manifolds (PNDM, PLMS | ICLR2022)

Code for CRATE (Coding RAte reduction TransformEr).

PyTorch implementation of MAR+DiffLoss https://arxiv.org/abs/2406.11838

[ICML 2024] Break the Sequential Dependency of LLM Inference Using Lookahead Decoding

800,000 step-level correctness labels on LLM solutions to MATH problems

Fast, memory-efficient, scalable optimization of deep learning with differential privacy

Implementation of Denoising Diffusion Probabilistic Model in Pytorch

Fully featured implementation of Routing Transformer

Sinkhorn Transformer - Practical implementation of Sparse Sinkhorn Attention

Transformer based on a variant of attention that is linear complexity in respect to sequence length

[ICML 2021 Oral] We show pure attention suffers rank collapse, and how different mechanisms combat it.