arXiv:2605.11599v1 Announce Type: new
Abstract: Fixed reasoning benchmarks evaluate canonical prompts, but semantically valid changes in presentation can still change model behavior. Studies of prompt variation can reveal such failures, but without au…
arXiv:2605.11428v1 Announce Type: new
Abstract: Exploratory analysis of high-dimensional data rarely stops at a single embedding. In practice, analysts rerun dimensionality reduction after changing preprocessing, subsets, or hyperparameters, and stand…
arXiv:2605.11554v1 Announce Type: new
Abstract: Task-agnostic structure proxies are often used to interpret why one pretraining corpus transfers better than another, but such explanations require the proxy to track the structure that matters for the d…