Learn AI

Sign In

by Amazon Research Publication

Amazon Research PublicationFollow

How Optimization and Forecasting Power Amazon Fulfillment

Deploying programmatic tool calling with pre-execution validation for production agentic systems
...
...
Amazon Research PublicationFollow

How Optimization and AI Power Amazon’s Fulfillment Network

On optical add/drop multiplexing architecture for hyperscale backbone networks
...
...
Amazon Research PublicationFollow

How Optimization and Forecasting Power Amazon Fulfillment

Beyond the harness: End-to-end optimization of context artifacts for enterprise Text-to-SQL
...
...
Amazon Research PublicationFollow

How to Evaluate LLM Agents Beyond Task Success

Beyond Task Success: An eight-metric tiered evaluation protocol for LLM agents over fragmented operational data
...
...
Amazon Research PublicationFollow

How Optimization and AI Power Amazon’s Delivery Network

MapScout: An agentic harness for map editing and geospatial data labeling
...
...
Amazon Research PublicationFollow

Off-Policy Ranking Evaluation: Rethinking Position Weights

Generalized position-based model: Rethinking position weights in ranking off-policy evaluation
...
...
Amazon Research PublicationFollow

The blind curator: How a biased judge silently disables skill retirement in self-evolving agents

The blind curator: How a biased judge silently disables skill retirement in self-evolving agents
...
...
Amazon Research PublicationFollow

How Optimization and Forecasting Power Amazon Fulfillment

Query-aware index pruning for retrieval under budget constraints
...
...
Amazon Research PublicationFollow

How Optimization and AI Power Amazon’s Fulfillment Network

Scalable conflation for maps data replay between heterogeneous geospatial data sources
...
...
Amazon Research PublicationFollow

How WARP Aligns RAG with Population Opinions

WARP: Wasserstein-Aligned RAG for population opinions
...
...
Amazon Research PublicationFollow

How to Build Realistic AI Twins of Mobile Users

AgenTwin: An end-to-end framework for building agentic twins of mobile users
...
...
Amazon Research PublicationFollow

How Sparse Attention Speeds Up Language Model Inference

SAS: Sparse attention synthesizer for efficient language model inference
...
...
Amazon Research PublicationFollow

How AI Turns BIM Requirements into Machine-Checkable IDS

Ishigaki-IDS-Bench: A benchmark for generating information delivery specification from BIM information requirements
...
...
Amazon Research PublicationFollow

How AI Learns to Check Evidence in Text and Images

Eliciting self-verification in multimodal reasoning agents with reinforcement learning
...
...
Amazon Research PublicationFollow

How Connectivity Shapes Attacks on Multi-Agent Recommenders

Attacking and defending multi-agent collaborative filtering systems through connectivity
...
...
Amazon Research PublicationFollow

How AI Finds Errors in Seismic Data Workflows

Analyzing seismic workflows with large language models: A case study on error detection
...
...
Amazon Research PublicationFollow

What Static Pruning Transfers Across Sparse Search Engines

Static pruning across sparse retrieval regimes: What transfers, what breaks, and what still helps
...
...
Amazon Research PublicationFollow

Predicting Trajectories from Historical Movement Patterns

Non-parametric spatiotemporal trajectory prediction via state-conditioned transition sampling
...
...
Amazon Research PublicationFollow

How Static Analysis Generates Least-Privilege IAM Policies

IAM policy autopilot: Static analysis for policy generation from application code
...
...
Amazon Research PublicationFollow

Why Pairwise Ranking Beats RL for Offline Explanation Choice

Pairwise ranking outperforms single-action RL for offline explanation selection: A practical lesson
...
...