Skip to content

cleanlab/tlm

Repository files navigation

Trustworthy Language Model (TLM)

The Trustworthy Language Model scores the trustworthiness of outputs from any LLM in real-time.

Automatically detect hallucinated/incorrect responses in: Q&A (RAG), Chatbots, Agents, Structured Outputs, Data Extraction, Tool Calling, Classification/Tagging, Data Labeling, and other LLM applications.

Use TLM to:

  • Guardrail AI mistakes before they are served to user
  • Escalate cases where AI is untrustworthy to humans
  • Discover incorrect LLM (or human) generated outputs in datasets/logs
  • Boost AI accuracy

Powered by uncertainty estimation techniques, TLM works out of the box, and does not require:
data preparation/labeling work or custom model training/serving infrastructure.

Learn more and see precision/recall benchmarks with frontier models (from OpenAI, Anthropic, Google, etc):
Blog, Research Paper

Usage

See notebooks for Jupyter notebooks with example usage.