Construct · 1 fields · 1 citations

unfaithful chain-of-thought: a standalone construct

One name. One construct. A landmark on the map.

● anchor · a standalone construct Nearest construct by meaning·not indexed

A single standalone construct; nearby constructs are located against it in the Collider. Tap a field for its definition, unit, and measures.

Natural language processing

a model's stated step-by-step reasoning misrepresents the real reasons for its answer, for example hiding the influence of biased features.

In natural language processing, unfaithful chain-of-thought refers to a model's stated step-by-step reasoning misrepresents the real reasons for its answer, for example hiding the influence of biased features. This meaning comes from Turpin (2023) in arXiv.[1]

Unit
Recovery means
Measured by

Run it through the Collider, or take the citations with you.

See how unfaithful chain-of-thought sits against neighbouring constructs in the Collider, or export every reference on this page in one click.

Open in Collider Export BibTeX Report this entry

Nearby constructs

influence5 fields · 5 citationsjingle · one word, five constructs model3 fields · 3 citationsjingle · one word, three constructs natural3 fields · 7 citationsjingle · one word, three constructs

References

Turpin, M., Michael, J., Perez, E., & Bowman, S. R. (2023). Language Models Don't Always Say What They Think: Unfaithful Explanations in Chain-of-Thought Prompting. arXiv. https://doi.org/10.48550/arxiv.2305.04388

Cited on this page?

Confirm your construct, or correct us.

If your work is cited behind one of these entries, you know better than we do whether we captured the meaning right. Confirm it, suggest an edit, or dispute it, and your name goes on the entry as an author who has checked it, linked to your ORCID.

Claim your construct