Amazon Science ORG

Chronos: Pretrained Models for Time Series Forecasting

Official implementation for "Multimodal Chain-of-Thought Reasoning in Language Models" (stay tuned and more will be updated)

Official implementation for "Automatic Chain of Thought Prompting in Large Language Models" (stay tuned & more will be updated)

No description.

RAGChecker: A Fine-grained Framework For Diagnosing RAG

SiamMOT: Siamese Multi-Object Tracking

Official implementation of Earthformer

RefChecker provides automatic checking pipeline and benchmark dataset for detecting fine-grained hallucinations generated by Large Language Models.

BigDetection: A Large-scale Benchmark for Improved Object Detector Pre-training

Shopping Queries Dataset: A Large-Scale ESCI Benchmark for Improving Product Search

No description.

Pytorch implementation of Supporting Clustering with Contrastive Learning, NAACL 2021

Repository for Fraud Dataset Benchmark

Official implementation for the paper "Prompt Pre-Training with Over Twenty-Thousand Classes for Open-Vocabulary Visual Recognition"

Official PyTorch implementation of TSDiff models presented in the NeurIPS 2023 paper "Predict, Refine, Synthesize: Self-Guiding Diffusion Models for Probabilistic Time Series Forecasting"

Project for <SPot-the-Difference Self-Supervised Pre-training for Anomaly Detection and Segmentation> (ECCV 2022)

ReFinED is an efficient and accurate entity linking (EL) system.

No description.

Official Implementations of "Mixed-Type Tabular Data Synthesis with Score-based Diffusion in Latent Space""

AuctionGym is a simulation environment that enables reproducible evaluation of bandit and reinforcement learning methods for online advertising auctions.

CrossCodeEval: A Diverse and Multilingual Benchmark for Cross-File Code Completion (NeurIPS 2023)

JuLS is a Julia Local Search solver that combines Constraint Based Local Search (CBLS) and Constraint Programming (CP)

No description.

No description.

[NeurIPS 2021 Spotlight] Official implementation of Long Short-Term Transformer for Online Action Detection

This package provides a pythorch implementation of "GAN-Control: Explicitly Controllable GANs", ICCV 2021.

Structured Prediction as Translation between Augmented Natural Languages

No description.

Official implementation for Training LLMs with MXFP4

CodeSage: Code Representation Learning At Scale (ICLR 2024)

Dataset from the paper "Mintaka: A Complex, Natural, and Multilingual Dataset for End-to-End Question Answering" (COLING 2022)

No description.

No description.

PyTorch implementation of EMAN for self-supervised and semi-supervised learning: https://arxiv.org/abs/2101.08482

Cyber-Zero: Training Cybersecurity Agents Without Runtime

No description.

This is an official implementation of TubeR: Tubelet Transformer for Video Action Detection

SWE-PolyBench: A multi-language benchmark for repository level evaluation of coding agents

Code repo for the ICML 2024 paper "Automated Evaluation of Retrieval-Augmented Language Models with Task-Specific Exam Generation"

Dataset associated with "BOLD: Dataset and Metrics for Measuring Biases in Open-Ended Language Generation" paper

No description.

No description.

Official implementation of the CVPR'24 paper [Adaptive Slot Attention: Object Discovery with Dynamic Slot Number]

Redset is a dataset containing three months worth of user query metadata that ran on a selected sample of instances in the Amazon Redshift fleet. We provide query metadata for 200 provisioned and serverless instances each.

No description.

PyTorch implementation of Omni-DETR for omni-supervised object detection: https://arxiv.org/abs/2203.16089

No description.

Repoformer: Selective Retrieval for Repository-Level Code Completion (ICML 2024)

No description.

PyTorch implementation of Semi-supervised Vision Transformers

CrossCLR: Cross-modal Contrastive Learning For Multi-modal Video Representations, ICCV 2021

Official implementation for the paper "A Cheaper and Better Diffusion Language Model with Soft-Masked Noise"

AgentOccam: A Simple Yet Strong Baseline for LLM-Based Web Agents

CaML: Carbon Footprinting of Household Products with Zero-Shot Semantic Text Similarity

Releasing code for "ReCode: Robustness Evaluation of Code Generation Models"

Official Implementation for ICCV'23 paper Coarse-to-Fine Amodal Segmentation with Shape Prior (C2F-Seg).

Training Language Model Agents to Find Vulnerabilities with CTF-Dojo

Datasets and code for results presented in the ProbConserv paper

Implementation of the paper "FactGraph: Evaluating Factuality in Summarization with Semantic Graph Representations (NAACL 2022)"

Code associated with the "Data Augmentation using Pre-trained Transformer Models" paper

DQ-BART: Efficient Sequence-to-Sequence Model via Joint Distillation and Quantization (ACL 2022)

Causal Inference in Python

Datasets and code for results presented in the BOON paper

[ICML 2025] Official implementation of Spherical Diffusion Policy: A SE(3) Equivariant Visuomotor Policy with Spherical Fourier Representation

Code for AAAI'24 paper "Graph Neural Prompting with Large Language Models".

No description.

No description.

No description.

Code for "Task-Agnostic Continual RL: In Praise of a Simple Baseline"

No description.

Code for the ACL 2023 paper: "Rethinking the Role of Scale for In-Context Learning: An Interpretability-based Case Study at 66 Billion Scale"

No description.

PipeRAG: Fast Retrieval-Augmented Generation via Algorithm-System Co-design (KDD 2025)

Implementation of the paper "Efficiently Summarizing Text and Graph Encodings of Multi-Document Clusters (NAACL 2021)"

No description.

No description.

No description.

Official implementation of Adaptive Feature Transfer (AFT)

No description.

LC-PLM: long-context protein language model based on BiMamba-S architecture

TurboFuzzLLM: Turbocharging Mutation-based Fuzzing for Effectively Jailbreaking Large Language Models in Practice

No description.

No description.

Scaling AI made stupid. Accelerate your AI experiments to multi-CPU/multi-node with 3 lines of code.

MO-LightGBM is a gradient boosting framework based on decision tree algorithms, used for Multi-objective learning to rank tasks.

An easy-to-configure and extensible veRL extension for agent RL training with skill co-evolution.

No description.

No description.

No description.

Enhancing contextual understanding in large language models through contrastive decoding

Synthesizing realistic and diverse text-datasets from augmented LLMs

AdaICL: Which Examples to Annotate of In-Context Learning? Towards Effective and Efficient Selection

No description.

No description.

This package provides a pythorch implementation of "FPGAN-Control: A Controllable Fingerprint Generator for Training with Synthetic Data", WACV 2024.

No description.

Code for paper "Factual Confidence of LLMs: on Reliability and Robustness of Current Estimators"

Official implementation of Spectro-Riemannian Graph Neural Networks (ICLR 2025)

The h3-indexer is an open source package for indexing geospatial data using PySpark, Apache Sedona and the H3 hierarchical spatial indexing system. The h3-indexer maps any number of vector-type geospatial data sets to H3 grids for efficient spatial analysis and querying.

Pre-training Intent-Aware Encoders for Zero- and Few-Shot Intent Classification

No description.

No description.

Code for the paper: Comparing and Contrasting Deep Learning Weather Prediction Backbones on Navier-Stokes and Atmospheric Dynamics

No description.

No description.

No description.

[ACL 2025] Learning to Reason Over Time: Timeline Self-Reflection for Improved Temporal Reasoning in Language Models

Deep Multi-task Learning to Rank

Official implementation of OpenTab (ICLR2024)

No description.

No description.

No description.

No description.

Implementation of the paper: Beyond Correlation: The impact of human uncertainty in measuring the effectiveness of automatic evaluation and LLM-as-a-judge

SafeSearch: a reinforcement learning framework that jointly optimizes safety and utility in LLM-based search agents.

No description.

No description.

No description.

No description.

Official implementation of "Demystifying Transition Matching: When and Why It Can Beat Flow Matching" (AISTATS 2026). Code for image and video generation using Transition Matching.

Dataset of GNSS location trajectories from both low-cost high-precision GNSS system and ground-truth reference GNSS system

Code the ICML 2024 paper: "Variance-reduced Zeroth-Order Methods for Fine-Tuning Language Models"

No description.

Implemental for the paper "Large Language Model Critics for Execution-Free Evaluation of Code Changes"

Algorithms to solve lattice problems in Julia

No description.

No description.

No description.

No description.

No description.

No description.

No description.

No description.

Expert hallucination labels from the ACI-bench dataset

Official repository for CiteEval: Principle-Driven Citation Evaluation for Source Attribution

No description.

WRAVAL helps in evaluating LLMs for writing assistant tasks like summarization, professional tone, witty tone, etc.

No description.

No description.

Code to support the publication entitled "Context-aware Multi-Property Antibody Predictor: a Novel Framework Integrating Text and Protein Language Models"

No description.

No description.

Official Implementation of ProbHardE2E: End-to-End Probabilistic Framework for Learning with Hard Constraints

No description.

Validate your causal models!

Evolutionary Generation of Multi-Agent Systems; Yuntong Hu, Yuting Zhang, Matthew Trager, Yi Zhang, Shuo Yang, Wei Xia, Stefano Soatto, ICML 2026

Official implementation of Operator-ProbConserv: OOD UQ for Neural Operators

No description.

Exploring Fine-Tuning for In-Context Retrieval and Efficient KV-Caching in Long-Context Language Models

Implementation of the EMNLP'24 paper: Salient Information Prompting to Steer Content in Prompt-based Abstractive Summarization

Python package for stratifying, sampling, and estimating model performance with fewer annotations.

No description.

No description.

No description.

No description.

No description.

Code for the paper RMIR: A Benchmark Dataset for Reasoning-Intensive Multimodal Image Retrieval.

Code repository for PersonaLens paper (ACL 2025).

CSR-Bench: Benchmarking LLM Agents in Deployment of Computer Science Research Repositories

Code the ICML 2024 paper: "MADA: Meta-Adaptive Optimizers through hyper-gradient Descent"

Implementation of DualKV: Shared-Prompt Flash Attention for Efficient RL Training with Large Rollouts and Long Contexts

No description.

No description.

No description.

No description.

No description.

No description.

RAGferee | Building Contextual Reward Models for Retrieval-Augmented Generation

No description.

Test optimized LLMs for degraded accuracy

No description.

No description.

No description.

No description.

Benchmark test sets for real-world noise phenomena in goal-directed conversations in English.

Experimental results capturing the limits on annotation noise under which MV can aggregate labels optimally.

🔬 Replication Package for the paper: "Benchmarking Multilingual Temporal Reasoning in LLMs: The Temporal Reasoning Dataset"

Understanding and Improving the Information Preservation in Prompt Compression for LLMs

No description.

[ACL 2026 main] Long-Context Hallucination Detection Benchmark: Rethinking Evaluation for LLM Hallucination Detection: A Desiderata, A New RAG-based Benchmark, New Insights

No description.

Official Implementation of Understanding Transformers for Time Series: Rank Structure, Flow-of-Ranks, and Compressibility

Official Implementation of Understanding the Implicit Biases of Design Choices for Time Series Foundation Models

No description.

No description.

CompAgent: An Agentic Framework for Visual Compliance Verification

No description.

No description.

No description.

No description.

No description.

Code for the AISTATS 2024 paper: "Krylov Cubic Regularized Newton: A Subspace Second-Order Method with Dimension-Free Convergence Rate".

UniqSketch: sensitive and resource-efficient strain-level detection in metagenomes

No description.

No description.

No description.

No description.

No description.

No description.

Code for ACL 2026 Paper "SWAN: Semantic Watermarking with Abstract Meaning Representation"