Benchmarks › DecodingTrustToxicity

DecodingTrustToxicity

Benchmark evaluates the toxicity of sentence completions.

Full title

DecodingTrustToxicity

Tag

decodingtrust-toxicity (use this to limit an assistant's search to this benchmark)

Reference

decodingtrust.github.io

BenchRisk

Scored 2025-04-01 · 39 mitigations adopted, 147 absent · score page · source entry

Documents available to assistants

Summary from the reference

DecodingTrust aims at providing a thorough assessment of trustworthiness in GPT models. This research endeavor is designed to help researchers and practitioners better understand the capabilities, limitations, and potential risks involved in deploying these state-of-the-art Large Language Models (LLMs). This project is organized around the following eight primary perspectives of trustworthiness, including: * Toxicity * Stereotype and bias * Adversarial robustness * Out-of-Distribution Robustness * Privacy * Robustness to Adversarial Demonstrations * Machine Ethics * Fairness

Ask an assistant about it

Once your assistant is connected, try a question such as “Using the benchmark sources, how is DecodingTrustToxicity scored?”

Developers can restrict a search to this benchmark with the filter {"benchmark": "decodingtrust-toxicity"} on the MCP endpoint https://ns-f1fc49e3-9bca-4aa2-b5a1-ef5dbc18588f.search.ai.cloudflare.com/mcp.