Papers/2609.15990
🧪 Test?View on arXiv

Few-Shot Degradation Is Not What It Seems: Behavioral Evidence, Representation Analysis, and a Random-Text Control Across 12 Models, 2 Tasks, and 2 Architectures

Not provided in the content

few-shot learninglanguage modelsrepresentation analysisprompt engineering
2609.15990
Builder Relevance
70%
1h ago

Abstract

The paper investigates the unexpected degradation of language models under few-shot prompting, revealing that the effect is task-dependent and proposing a method to isolate the impact of prompt length from content.

Reality Card

Core Claim

The proposed metric 'content delta' effectively predicts the performance of few-shot prompting, showing that models benefiting from demonstration content restructure their representations more.

Method / Result

Content delta correlates with performance (rho = +0.65, p = 0.043), while raw shift does not (r = 0.20).

Limitations

The study's findings may be limited by the specific tasks and models evaluated, which may not generalize across all applications.

Paper to code

Verified implementation resources so builders can test the paper’s claims instead of stopping at the abstract.

No verified implementation link has been attached yet. AIBuzzHub will keep this panel separate from unverified search results.
← Back to all papers