Launch Week 3: Five days of launches
Confident AI
Docs
Pricing
Products
LLM Evaluation
LLM Evaluation

Benchmark LLM systems with research-backed metrics.

LLM Observability
LLM Observability

Trace, monitor, and alert on production LLM systems.

AI Red Teaming
AI Red Teaming

Stress-test LLM apps against adversarial attacks.

AI Governance
AI Governance

Enforce AI standards and controls across teams.

Book a DemoSign Up
Book a DemoSign Up

Step Efficiency

Step Efficiency is a single-turn metric to determine if your agent took steps it did not need to take

Overview

The step efficiency metric is a single-turn metric that uses LLM-as-a-judge to assess whether your LLM agent completed its task without taking steps it did not need to take.

How Is It Calculated?

The step efficiency metric uses an LLM to extract the task from the trace, then uses the same LLM to score the execution steps your agent took against that task, penalizing every action that was not strictly required to finish it.


Step Efficiency=Alignment Score(Task,Execution Steps)\text{Step Efficiency} = \text{Alignment Score}(\text{Task}, \text{Execution Steps})

The final score is the alignment of the execution steps with the task as extracted from the trace.

Usage

To run the step efficiency metric on Confident AI, add it to a single-turn metric collection. This will allow you to use step efficiency metric for:

  • Single-turn E2E testing
  • Online and offline evals for traces

Last updated on

Built byConfident AI