Construct · 1 fields · 1 citations

superficial alignment hypothesis: 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

the claim that a model learns almost all knowledge in pretraining and alignment mainly teaches the format and style of responses.

In natural language processing, superficial alignment hypothesis refers to the claim that a model learns almost all knowledge in pretraining and alignment mainly teaches the format and style of responses. This meaning comes from Zhou (2023) in arXiv.[1]

Unit
Recovery means
Measured by

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

See how superficial alignment hypothesis 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

alignment6 fields · 12 citationsjingle · one name, six constructs hypothesis2 fields · 2 citationsjingle · one word, two constructs knowledge6 fields · 12 citationsjingle · one name, six constructs

References

Zhou, C., Liu, P., Xu, P., Iyer, S., Sun, J., Mao, Y., Ma, X., Efrat, A., Yu, P., Yu, L., Zhang, S., Ghosh, G., Lewis, M., Zettlemoyer, L., & Levy, O. (2023). LIMA: Less Is More for Alignment. arXiv. https://doi.org/10.48550/arxiv.2305.11206

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