expertise needed
evaluation design
Requires skill in designing test suites, benchmarks, adversarial scenarios and rating or annotation protocols to judge model behaviour, including building ground-truth sets and deciding what counts as a pass.
24 techniques
| Goals | Models | Data Types | Description | |||
|---|---|---|---|---|---|---|
| Causal Discrimination Testing | Testing | Architecture/model Agnostic Paradigm/supervised +1 | Tabular Text | Causal discrimination testing checks whether a system gives different outputs to otherwise identical inputs because of a... | ||
| Safety Envelope Testing | Testing | Architecture/model Agnostic Requirements/black Box | Any | Safety envelope testing checks how an AI system performs at and beyond the edges of the conditions it was designed for,... | ||
| Prompt Robustness Testing | Testing | Architecture/neural Networks/transformer/llm Paradigm/generative +1 | Text | Prompt robustness testing measures how much a language model's answers, or its performance on a task, change when the... | ||
| Agent Goal Misalignment Testing | Testing | Architecture/model Agnostic Architecture/neural Networks/transformer/llm +3 | Text | Agent goal misalignment testing checks whether an AI agent takes unsanctioned means to reach a goal it has been given.... | ||
| Chain-of-Thought Faithfulness Evaluation | Testing | Architecture/neural Networks/transformer/llm Paradigm/generative +1 | Text | Chain-of-thought faithfulness evaluation tests whether the step-by-step reasoning a language model writes out is what... | ||
| Constitutional AI Evaluation | Testing | Architecture/neural Networks/transformer/llm Requirements/white Box +1 | Text | Constitutional AI evaluation checks whether a language model follows the written principles, or 'constitution', that it... | ||
| Few-Shot Fairness Evaluation | Testing | Architecture/neural Networks/transformer/llm Paradigm/generative +1 | Text | Few-shot fairness evaluation tests whether the worked examples placed in a prompt change how fairly a language model... | ||
| Jailbreak Resistance Testing | Testing | Architecture/neural Networks/transformer/llm Requirements/black Box | Text | Jailbreak resistance testing checks whether a language model or chatbot can be talked out of its safety rules. Testers... | ||
| Knowledge Boundary Probing | Algorithmic | Architecture/neural Networks/transformer/llm Requirements/black Box | Text | Knowledge boundary probing maps what a language model can and cannot reliably answer. A probe set is built from... | ||
| Cross-Modal Consistency Scoring | Metric | Architecture/model Agnostic Requirements/black Box | Image Text | Cross-modal consistency scoring checks whether content in one modality matches content in another, such as an image and... | ||
| Prompt Injection Testing | Testing | Architecture/neural Networks/transformer/llm Paradigm/generative +1 | Text | Prompt injection testing checks whether an AI system can be made to follow instructions its builders did not intend. In... | ||
| Retrieval-Augmented Generation Evaluation | Testing | Architecture/neural Networks/transformer/llm Paradigm/generative +1 | Text | A retrieval-augmented generation (RAG) system answers a question by retrieving passages from a document collection and... | ||
| Reward Hacking Detection | Testing | Architecture/model Agnostic Architecture/neural Networks/transformer/llm +2 | Any | Reward hacking is when an agent trained to maximise a proxy reward (a score that only approximates what its designers... | ||
| Toxicity Detection | Testing | Architecture/model Agnostic Architecture/neural Networks/transformer/llm +2 | Text | Toxicity detection scores a model's outputs with a classifier trained to recognise abusive language, such as insults,... | ||
| Scenario-Based Simulation Testing | Testing | Architecture/model Agnostic Requirements/black Box | Any | Scenario-based simulation testing checks an automated system, most often an automated driving system, by running it... | ||
| PII Detection and Redaction in Training Text | Algorithmic | Architecture/model Agnostic Architecture/neural Networks/transformer/llm +1 | Text | PII detection and redaction finds personally identifiable information (PII), such as names, email addresses, phone... | ||
| Dangerous Capability Evaluation | Testing | Architecture/neural Networks/transformer/llm Paradigm/generative +1 | Text | Dangerous capability evaluation measures whether a frontier AI model can carry out tasks that could enable severe harm,... | ||
| SMACTR Internal Algorithmic Audit | Process | Architecture/model Agnostic Requirements/black Box | Any | SMACTR is an internal audit procedure for AI systems, proposed by Raji and colleagues in 2020, in which an audit team... | ||
| Metamorphic Testing | Testing | Architecture/model Agnostic Requirements/black Box | Any | Metamorphic testing checks a model or AI system without needing the correct output for each test input, by checking that... | ||
| Deletion and Insertion Curves | Testing | Architecture/model Agnostic Requirements/black Box | Image Text Time-series | Deletion and insertion curves test whether a feature-attribution explanation, such as a heatmap over an image or... |
Rows per page
Page 1 of 2