Bibliographic record
Abstract
Compressed sensing (CS) is a paradigm in which a structured high-dimensional signal may be recovered from random, under-determined, corrupted linear measurements. LASSO programs are effective for solving CS problems due to their proven ability to leverage underlying signal structure. Three popular LASSO programs are equivalent in a sense and sometimes used interchangeably. Tuned by a governing parameter, each admits an optimal parameter choice yielding minimax order-optimal error. CS is well-studied, though theory for LASSO programs typically concerns this optimally tuned setting. However, the optimal parameter value for a LASSO program depends on properties of the data and is typically unknown in practical settings. Performance in empirical problems thus hinges on a program's parameter sensitivity: it is desirable that small variation about the optimal parameter choice begets small variation about the optimal risk. We examine the risk of three LASSO programs as a function of their governing parameters and further demonstrate that their parameter sensitivity can differ for the same data, thereby informing the selection of a LASSO program in practice. We prove a gauge-constrained program admits asymptotic cusp-like behaviour of its risk in the limiting low-noise regime; a residual-constrained program has asymptotically suboptimal risk for very sparse vectors (i.e., for any fixed parameter choice, the ratio of the risk to the optimally achievable risk grows unbounded). These results contrast observations about an unconstrained program with sufficiently large parameter. Our theory is supported with extensive numerical simulations, demonstrating the parameter sensitivity phenomenon for even modest dimensional parameters. We first analyze these risks for proximal denoising (PD), in which one directly observes signal plus noise (i.e., the measurement matrix is identity). There, we further reveal a data regime in which the unconstrained PD risk can be asymptotically suboptimal. We also show how our theory extends to analyze generalized LASSO programs for generalized CS. Finally, we extend a keystone of our theoretical analysis, the projection lemma. We generalize the result to an arbitrary Hilbert space and extend it from scaled projections to proximal mappings of a dilated gauge. We discuss applications and possible directions for these results.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.000 | 0.000 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.000 | 0.000 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.000 | 0.000 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.001 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one teacher head, not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".