FeedSourcesModels
PrivacyImpressum
⌘K
AS

Amazon Science Publications

128 articles total

Go to source

Amazon is a great place to practice science and have real business impact with publications. Amazon scientists continue to publish papers, teach, and engage with the worldwide research community.

Go to source
  •  Failing agents leave traces: Telemetry signatures for multi-tier agent evaluation
  •  SynthAVE: Scalable synthetic labeling for e-commerce with LLM-arena validation
  •  ESPO: Error-structured prompt optimization via diagnose, diversify, and stabilize
  •  Chart-RL: Policy optimization reinforcement learning for enhanced visual reasoning in chart question answering with Vision Language Models
  •  Validation-gated architecture search for self-organizing multi-agent orchestration
  •  Characterizing and provisioning the environment bubble in agentic-RL LLM rollout
  •  Efficient enumeration of cloud access permissions
  •  From trace entropy to coordination control: Diagnosing and intervening in multi-Agent LLM systems
  •  Bee private compute: Privacy-preserving intelligence for Bee
  •  DI-Bench: Systematically generating in-domain benchmarks for data intelligence agents
  •  PersonaJudge: Simulating individual human preference judgments with evaluator-specific demonstration data
  •  Deploying programmatic tool calling with pre-execution validation for production agentic systems
  •  TRACE : Traceable root-cause analysis with calibrated evidence-grounded agents
  •  On optical add/drop multiplexing architecture for hyperscale backbone networks
  •  ChainAlign: Cross-lingual sentence alignment via sparse co-linear chaining
  •  Data-driven personas for survey simulation: Insights into simulation alignment across data-access regimes
  •  RetroAgent: Harnessing LLMs to search over structured memory for agentic retrosynthesis planning
  •  Beyond the harness: End-to-end optimization of context artifacts for enterprise Text-to-SQL
  •  Beyond Task Success: An Eight-Metric Tiered Evaluation Protocol for LLM Agents over Fragmented Operational Data
  •  MapScout: An agentic harness for map editing and geospatial data labeling