Skip to main content

Measure how much an LLM’s answers drift to detect hallucination risk and improve prompts for more reliable results.

Project description

hallu-assess

Measure how much an LLM’s answers drift to detect hallucination risk and improve prompts for more reliable outputs.

What it does

hallu-assess evaluates hallucination risk by:

  • running the same prompt multiple times
  • embedding each response
  • measuring semantic deviation across outputs

High deviation → higher hallucination risk.

Installation

pip install hallu-assess

Detailed Documentation - WIP

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

hallu_assess-0.1.0.tar.gz (6.3 kB view details)

Uploaded Source

Built Distribution

If you're not sure about the file name format, learn more about wheel file names.

hallu_assess-0.1.0-py3-none-any.whl (8.2 kB view details)

Uploaded Python 3

File details

Details for the file hallu_assess-0.1.0.tar.gz.

File metadata

  • Download URL: hallu_assess-0.1.0.tar.gz
  • Upload date:
  • Size: 6.3 kB
  • Tags: Source
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/6.2.0 CPython/3.12.6

File hashes

Hashes for hallu_assess-0.1.0.tar.gz
Algorithm Hash digest
SHA256 c80aa64783c5f8e0730b9cca02c9d96cb32fc44d8f9d4f60c42a40ac743f9302
MD5 001d0ef851e8360c0744ed6f9d45aa9f
BLAKE2b-256 584432b858edf57d5dc9f48c16bf013c1388498a4f00970d203e554cc377f059

See more details on using hashes here.

File details

Details for the file hallu_assess-0.1.0-py3-none-any.whl.

File metadata

  • Download URL: hallu_assess-0.1.0-py3-none-any.whl
  • Upload date:
  • Size: 8.2 kB
  • Tags: Python 3
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/6.2.0 CPython/3.12.6

File hashes

Hashes for hallu_assess-0.1.0-py3-none-any.whl
Algorithm Hash digest
SHA256 fef99d469a80b5f51d4ce4b95f4509e0ea7b9980a3725edf8ad9c3aa5486636c
MD5 408154b657007bbf6f3c08de0628f296
BLAKE2b-256 05bc782fdc6c54f757ef6851df391805540ac65604aa0e4f81f5d0e493690e4c

See more details on using hashes here.

Supported by

AWS Cloud computing and Security Sponsor Datadog Monitoring Depot Continuous Integration Fastly CDN Google Download Analytics Pingdom Monitoring Sentry Error logging StatusPage Status page